ARRT MRI Registry Exam Prep: Blueprint Map, Practice Questions, and Study Plan
The ARRT MRI exam has 200 scored questions, and you pass with a scaled score of 75. The current 2026 handbooks list different pilot-item counts and timing by eligibility pathway: primary candidates get 230 total items with 230 minutes of test time; postprimary candidates get 220 total items with 210 minutes of test time. Here's where the 200 scored questions come from, then 24 original practice questions to start on right now. ARRT content specifications, effective Feb. 1, 2025 · Last verified October 27, 2026
Where your 200 scored questions come from
| Category | Scored questions | Share of your score |
|---|---|---|
| Image Production | 106 | 53% |
| Procedures | 57 | 28.5% |
| Safety | 21 | 10.5% |
| Patient Care | 16 | 8% |
More than half your score is Image Production: physics, parameters, sequences, and artifacts. Give it your biggest block of study time. The counts are ARRT's; the percentages are simple math on those counts. The full subcategory breakdown is further down.
Start here:
- Work the 24 practice questions below. At about a minute each, that's roughly 25 minutes.
- Log every miss by subcategory and by why you missed it (the four miss types are under the set).
- Run the 6-week plan, front-loading whatever you missed most.
Practice questions
24 original, unofficial questions written for this page. They aren't ARRT questions. The mix roughly follows ARRT's weights: 12 Image Production, 7 Procedures, 3 Safety, 2 Patient Care. Each answer stays hidden until you reveal it, and each explanation covers why the other choices are wrong.
Question 1 · MR-IP-01 · Image Production: Physical Principles of Image Formation
What is the precessional (Larmor) frequency of hydrogen protons in a 3.0 T magnet?
Reveal answer and explanation for Question 1 (MR-IP-01)
Answer: C. 127.74 MHz
The Larmor equation says frequency = gyromagnetic ratio × field strength. For hydrogen the gyromagnetic ratio is 42.58 MHz/T, so 42.58 × 3.0 = 127.74 MHz.
- A is the gyromagnetic ratio itself, which is the frequency at 1.0 T.
- B is the frequency at 1.5 T.
- D doubles the answer again, which would mean a 6.0 T magnet.
Review: Formulas worth memorizing · Sources: Hornak, The Basics of MRI, problem 3-1 (Rochester Institute of Technology)
Source access: The RIT answer page cited above was unavailable during this check. NIST's CODATA value for the proton gyromagnetic ratio in MHz/T rounds to the 42.58 MHz/T used in the worked answer.
Question 2 · MR-SA-01 · Safety: MRI Screening and Safety
During screening, a patient shows you an implant card stating the device is "MR Conditional." What should happen next?
Reveal answer and explanation for Question 2 (MR-SA-01)
Answer: B
"MR Conditional" means the device has been shown to pose no known hazard only under the specific conditions in its labeling. Your job is to identify the device precisely and confirm those conditions can be met. ARRT's Safety outline lists exactly this sequence: document the device, verify its labeling, and identify its device-specific parameters.
- A describes MR Safe, a different label. Passive metallic implants can't be MR Safe at all; they're labeled MR Conditional or MR Unsafe.
- C is wrong because conditional devices are routinely scanned once their conditions are confirmed.
- D is wrong because conditions apply at the field strengths the device was tested at. ACR's current manual notes that some implants heat more at 1.5 T than at 3 T, so a 3 T label can't be assumed safe at 1.5 T.
Review: Safety and Patient Care must-knows · Sources: ARRT MRI Content Specifications, effective Feb. 1, 2025; RSNA 2025 education exhibit on MR safety labeling (ASTM F2503); ACR, What's New in the Manual on MR Safety (March 2026)
Question 3 · MR-IP-02 · Image Production: Sequence Parameters and Options
A fast spin echo sequence uses TR 4000 ms, phase matrix 256, NEX 2, and echo train length (ETL) 16. What is the scan time?
Reveal answer and explanation for Question 3 (MR-IP-02)
Answer: C. 2 minutes 8 seconds
FSE scan time = (TR × phase matrix × NEX) ÷ ETL. Here that's (4000 × 256 × 2) ÷ 16 = 128,000 ms = 128 seconds = 2 minutes 8 seconds.
- A forgets to divide by the ETL (2,048,000 ms, about 34 minutes). That's the conventional spin echo time for the same settings.
- B is what you'd get with an ETL of 8.
- D forgets to multiply by NEX.
Review: Parameter tradeoffs · Sources: Korosec, Fast Spin Echo Imaging, ISMRM 2009 syllabus
Question 4 · MR-PC-01 · Patient Care: Patient Interactions and Management
An outpatient scheduled for contrast-enhanced MRI has an eGFR of 24 mL/min/1.73 m². Which response best reflects current American College of Radiology (ACR) guidance?
Reveal answer and explanation for Question 4 (MR-PC-01)
Answer: B
An eGFR below 30 mL/min/1.73 m² identifies severe chronic kidney disease in the ACR NSF guidance. For patients at risk for NSF, ACR strongly prefers Group II gadolinium-based contrast agents; in patients with CKD stage 4 or 5 who are not on chronic dialysis, Group I agents are contraindicated and a Group II agent should be used if a gadolinium-enhanced MRI is necessary. So this is a protocol-driven risk-benefit decision, not a reflexive yes or no.
- A ignores renal screening, which ARRT's Patient Care outline lists (BUN, creatinine, eGFR).
- C overstates the rule. It's a risk-benefit decision, not an absolute contraindication.
- D is wrong because ACR advises the lowest dose that answers the clinical question, not double dosing in at-risk patients.
Review: Safety and Patient Care must-knows · Sources: ACR Manual on Contrast Media; ACR and National Kidney Foundation consensus on gadolinium in kidney disease, Radiology (2021); ARRT MRI Content Specifications, effective Feb. 1, 2025
Question 5 · MR-IP-03 · Image Production: Physical Principles of Image Formation
On an axial abdomen with the patient's arms at the sides, the arms appear wrapped onto the opposite side of the image. The phase-encoding direction is right to left, and the field of view is narrower than the patient. What directly corrects this?
Reveal answer and explanation for Question 5 (MR-IP-03)
Answer: B
This is aliasing (wraparound): anatomy outside the field of view in the phase direction gets mapped back inside it. Frequency-direction wrap is handled by routine oversampling, so in-plane aliasing is practically a phase-direction problem. Widening the phase FOV or adding phase oversampling fixes it. Oversampling acquires extra phase lines, so expect a longer scan unless you compensate. Swapping the phase and frequency directions is another option.
- A reduces chemical shift artifact, not wrap.
- C reduces flow-related artifact.
- D changes contrast and signal, not wrap.
Review: Artifacts: cause, look, and fix · Sources: MRI Artifacts and How to Overcome Them, RadioGraphics (2022); AuntMinnieEurope, Know your artifacts to optimize MRI scans
Question 6 · MR-SA-02 · Safety: MRI Screening and Safety
A patient is positioned supine for a pelvis exam with the inner thighs touching and hands clasped on the abdomen. What's the best action before scanning?
Reveal answer and explanation for Question 6 (MR-SA-02)
Answer: B
Skin-to-skin contact can complete a conductive loop. The RF field induces current in that loop, and the small contact point can burn. Published prevention guidance calls for real pads (not just a sheet) at contact points, uncrossed limbs and unclasped hands, and padding between the patient and the bore.
- A: a sheet is too thin to break the loop reliably.
- C makes the loop worse.
- D: TE isn't the lever for RF heating, and no parameter change replaces correct positioning.
Review: Safety and Patient Care must-knows · Sources: Burn-prevention measures, Journal of Medical Radiation Sciences (2021), Table 1; Shah and Aran, A Review of MR Safety, Cureus (2023)
Question 7 · MR-IP-04 · Image Production: Sequence Parameters and Options
You double the NEX from 1 to 2 and change nothing else. What happens?
Reveal answer and explanation for Question 7 (MR-IP-04)
Answer: B. SNR increases about 1.41×; scan time doubles.
Signal adds directly with each average, but random noise adds only as the square root. So SNR rises by √2 (about 1.41×) when NEX doubles, while scan time, which is proportional to NEX, doubles. That's why extra averages are an expensive way to buy SNR.
- A and D overstate the SNR gain.
- C understates the time cost.
Review: Parameter tradeoffs · Sources: Graessner, The Signal-to-Noise Indicator, MAGNETOM Flash (Siemens Healthineers)
Question 8 · MR-PR-01 · Procedures: Neurological
In a brain-for-MS protocol, which sequence suppresses cerebrospinal fluid so periventricular white-matter lesions stand out?
Reveal answer and explanation for Question 8 (MR-PR-01)
Answer: B. FLAIR
FLAIR (fluid-attenuated inversion recovery) uses a long inversion time chosen to null CSF, which has a very long T1. Bright lesions beside the ventricles then stand out against dark fluid instead of blending into bright fluid.
- A: STIR uses a short TI to null tissue with a T1 near fat's, not CSF.
- C: T1 spin echo shows dark CSF but poor lesion conspicuity.
- D: T2* gradient echo highlights susceptibility effects such as blood products, not CSF-suppressed lesion contrast.
Review: Procedures: what to prioritize · Sources: MRIquestions.com, TI to null a tissue
Question 9 · MR-IP-05 · Image Production: Data Acquisition, Processing, and Storage
After gadolinium, the radiologist wants fat-suppressed images that show an enhancing lesion. Why is STIR a poor choice here?
Reveal answer and explanation for Question 9 (MR-IP-05)
Answer: B
STIR isn't fat-specific. Its short inversion time nulls anything with a T1 near fat's. Gadolinium shortens T1, so enhancing tissue can be suppressed along with fat, hiding exactly what you're trying to show. After contrast, use a fat-specific method (spectral fat saturation or Dixon) with T1 weighting.
- A: STIR is widely used at 1.5 T.
- C: STIR doesn't target water.
- D: contrast doesn't change SAR that way.
Review: Parameter tradeoffs · Sources: MRIquestions.com, STIR; Reeder, Fat-Water Imaging, ISMRM 2009 syllabus
Question 10 · MR-IP-06 · Image Production: Physical Principles of Image Formation
A bright band appears on one side of each kidney and a dark band on the other, along the frequency-encoding direction at the kidney–perirenal fat border. Which change reduces it?
Reveal answer and explanation for Question 10 (MR-IP-06)
Answer: A. Increase the receive bandwidth.
This is chemical shift artifact. Fat protons precess slightly slower than water protons, so the scanner maps fat a few pixels away from where it belongs along the frequency direction. A wider bandwidth per pixel makes that fixed frequency difference span fewer pixels. The tradeoff is lower SNR. Fat suppression or swapping the phase and frequency directions also help.
- B makes the shift larger.
- C and D change SNR or contrast but not the displacement.
Review: Artifacts: cause, look, and fix · Sources: AuntMinnieEurope, Know your artifacts to optimize MRI scans; MRIquestions.com, Chemical shift artifact
Question 11 · MR-PR-02 · Procedures: Neurological
You're running a 2D time-of-flight MRA of the carotid bifurcations with axial slices and want arteries bright and jugular veins dark. Where do you place the saturation band?
Reveal answer and explanation for Question 11 (MR-PR-02)
Answer: B. Superior to the slices
Time-of-flight relies on fresh, unsaturated blood flowing into the slices. In the neck, arterial blood flows up and jugular venous blood flows down, so a band above the slices saturates venous blood before it enters. For an MR venogram you'd reverse the logic and place the band below.
- A would saturate the arterial inflow you want to see.
- C doesn't sit in the path of either flow.
- D: without a band, venous signal contaminates the arterial images.
Review: Procedures: what to prioritize · Sources: Inflow (time-of-flight) effects and venous saturation, University of Michigan Deep Blue chapter
Source access: The linked legacy University of Michigan PDF could not be downloaded during this check. The university's repository metadata identifies the same chapter as Cheng and Haacke, Magnetic Resonance Angiography, Current Protocols in Magnetic Resonance Imaging (2002); the DOI links to its publisher record.
Question 12 · MR-IP-07 · Image Production: Physical Principles of Image Formation
A line of alternating bright and dark noise runs across every image at the same position. Afterward you find that the Zone IV door didn't close fully. What's the cause?
Reveal answer and explanation for Question 12 (MR-IP-07)
Answer: B
Zipper artifacts come from stray RF energy at a fixed frequency getting into the received signal. An open door or damaged door contacts break the RF shield and let outside signals in. ARRT's outline lists the zipper under radiofrequency artifacts.
- A causes distortion and signal loss near metal or air–tissue interfaces.
- C wraps anatomy across the image.
- D creates fine parallel lines along sharp high-contrast edges.
Review: Artifacts: cause, look, and fix · Sources: AuntMinnieEurope, Know your artifacts to optimize MRI scans; ARRT MRI Content Specifications, effective Feb. 1, 2025
Question 13 · MR-PR-03 · Procedures: Body
Which approach produces the cholangiogram-like images of an MRCP?
Reveal answer and explanation for Question 13 (MR-PR-03)
Answer: B
Bile and pancreatic secretions have long T2 values. With a very long TE, signal from most body tissue decays away while the slow-moving fluid in the ducts stays bright against a dark background. Single-shot or 3D fast spin echo T2 techniques are typical, often with MIP post-processing.
- A: fluid is dark on T1.
- C is built for diffusion contrast, not duct depiction.
- D depicts flowing blood, not static fluid.
Review: Procedures: what to prioritize · Sources: MR cholangiopancreatography technique review (PMC, 2012)
Question 14 · MR-IP-08 · Image Production: Sequence Parameters and Options
Which spin echo combination produces proton density (PD) weighting?
Reveal answer and explanation for Question 14 (MR-IP-08)
Answer: C. Long TR, short TE
A long TR lets most tissues recover their longitudinal magnetization, which minimizes T1 differences. A short TE samples before much T2 decay, which minimizes T2 differences. What's left is contrast driven mainly by proton density.
- A gives T1 weighting.
- B gives T2 weighting.
- D gives weak signal with muddled, unhelpful contrast.
Review: Parameter tradeoffs · Sources: UCLA M219 Lecture 15: Imaging Sequences II
Question 15 · MR-PR-04 · Procedures: Body
On chemical shift (in-phase and opposed-phase) imaging, an adrenal mass loses signal on the opposed-phase images compared with the in-phase images. What does this indicate?
Reveal answer and explanation for Question 15 (MR-PR-04)
Answer: B. Microscopic (intracellular) lipid mixed with water in the same voxels
When fat and water protons share a voxel, their signals add at the in-phase echo time and partly cancel at the opposed-phase echo time. A signal drop on opposed-phase images therefore points to intracellular lipid, the classic finding in lipid-rich adrenal adenomas. The same principle detects fatty liver.
- A: hemorrhage doesn't cause this fat–water cancellation.
- C: chemical shift imaging is usually done without contrast.
- D: motion causes ghosting, not a consistent opposed-phase signal drop.
Review: Procedures: what to prioritize · Sources: Chemical Shift MR Imaging of the Adrenal Gland, RadioGraphics (2016)
Question 16 · MR-PR-05 · Procedures: Neurological
A patient arrives three hours after stroke symptoms began. What appearance indicates acute infarction on diffusion imaging?
Reveal answer and explanation for Question 16 (MR-PR-05)
Answer: A. Bright on DWI and dark on the ADC map
Acute ischemia causes cell swelling that restricts water diffusion. Restricted diffusion is bright on DWI and dark (low ADC) on the ADC map. You need both, because DWI brightness alone can come from long T2 ("T2 shine-through").
- B is the T2 shine-through pattern: bright on DWI but normal or high ADC.
- C and D don't represent restricted diffusion. D is the expected pattern for free fluid such as CSF.
Review: Procedures: what to prioritize · Sources: Persistent DWI hyperintensity and ADC hypointensity in infarction (PMC); Time course of ADC and DWI signal in acute ischemic stroke (PMC)
Question 17 · MR-PR-06 · Procedures: Musculoskeletal
A patient with a metal tibial plate needs knee imaging near the hardware. Which approach minimizes the metal artifact?
Reveal answer and explanation for Question 17 (MR-PR-06)
Answer: B. Fast spin echo with an increased receive bandwidth
Spin echo sequences use 180° refocusing pulses that undo much of the dephasing that metal causes. A higher readout bandwidth reduces in-plane distortion near the hardware, at some SNR cost. Gradient echo has no refocusing pulse, so its signal voids are much larger.
- A worsens the artifact.
- C increases distortion.
- D is among the most susceptibility-sensitive techniques.
Review: Artifacts: cause, look, and fix · Sources: Metal-related artifacts: imaging techniques and challenges, ISMRM 2016; Hargreaves et al., Metal-induced artifacts in MRI, ISMRM 2011
Question 18 · MR-IP-09 · Image Production: Sequence Parameters and Options
On a phase-contrast flow study, the fastest flow in the center of a vessel appears to reverse direction. What's the fix?
Reveal answer and explanation for Question 18 (MR-IP-09)
Answer: B. Raise the VENC above the peak expected velocity.
VENC (velocity encoding) sets the highest velocity the sequence can represent. Flow faster than the VENC wraps around and looks like flow in the opposite direction, which is velocity aliasing. Set the VENC modestly above the fastest expected flow; one commonly cited rule of thumb is about 20–25% higher.
- A makes aliasing worse.
- C and D don't change the velocity range.
Review: Parameter tradeoffs · Sources: MRIquestions.com, Velocity encoding (VENC)
Question 19 · MR-PR-07 · Procedures: Musculoskeletal
An elbow imaged off-center with the arm at the patient's side shows patchy, failed spectral fat suppression. You need fluid-sensitive images without contrast. What's the best option?
Reveal answer and explanation for Question 19 (MR-PR-07)
Answer: A. Switch to STIR.
Spectral fat saturation depends on a uniform magnetic field, which is hard to achieve far from the magnet's center. STIR nulls fat by its T1 and is relatively insensitive to field inhomogeneity, so it suppresses fat more evenly off-center. The tradeoffs are lower SNR and a longer scan, and STIR isn't suitable after gadolinium (see Question 9). Dixon fat–water separation is another robust option where available.
- B adds SNR but doesn't fix uneven suppression.
- C isn't fluid-sensitive.
- D doesn't address the cause.
Review: Procedures: what to prioritize · Sources: MRIquestions.com, STIR
Question 20 · MR-IP-10 · Image Production: Sequence Parameters and Options
In a fast spin echo sequence, which echo mainly determines image contrast (the effective TE)?
Reveal answer and explanation for Question 20 (MR-IP-10)
Answer: B. The echo or echoes placed in the center of k-space
The center of k-space holds the low spatial frequencies that set overall signal and contrast. In FSE, phase encoding is ordered so the echo at the chosen effective TE fills the center lines. The edges of k-space mostly carry fine detail.
- A, C, and D name echo positions or an average that don't decide which data land in the center of k-space.
Review: Parameter tradeoffs · Sources: Korosec, Fast Spin Echo Imaging, ISMRM 2009 syllabus
Question 21 · MR-IP-11 · Image Production: Physical Principles of Image Formation
Why is a gradient echo sequence more sensitive to blood products, metal, and air–tissue interfaces than a spin echo sequence?
Reveal answer and explanation for Question 21 (MR-IP-11)
Answer: B
A spin echo's 180° pulse refocuses dephasing caused by static field inhomogeneity. Gradient echo forms its echo with a gradient reversal instead, so local field disturbances keep dephasing the spins. That T2* sensitivity is a problem near metal but useful for finding hemorrhage, and susceptibility-weighted imaging (SWI) exploits it on purpose.
- A: gradient echo typically uses shorter TRs.
- C: gradient echo usually uses smaller flip angles.
- D isn't true.
Review: Artifacts: cause, look, and fix · Sources: UCLA M219 Lecture 15: Imaging Sequences II; Hargreaves et al., Metal-induced artifacts in MRI, ISMRM 2011
Question 22 · MR-IP-12 · Image Production: Physical Principles of Image Formation
You keep the same receive bandwidth per pixel and move a protocol from 1.5 T to 3.0 T. What happens to chemical shift displacement?
Reveal answer and explanation for Question 22 (MR-IP-12)
Answer: C. It roughly doubles.
The fat–water difference is about 3.5 parts per million of the Larmor frequency, so in hertz it grows with field strength: roughly 224 Hz at 1.5 T and roughly 447 Hz at 3.0 T. At the same bandwidth per pixel, twice the frequency gap spans about twice as many pixels. That's one reason 3 T protocols often use higher bandwidths.
- A and B ignore that the frequency gap grows with field strength.
- D is wrong because the gap never disappears.
Review: Formulas worth memorizing · Sources: Reeder, Fat-Water Imaging, ISMRM 2009 syllabus; Hornak, The Basics of MRI, problem 3-1 (Rochester Institute of Technology)
Question 23 · MR-PC-02 · Patient Care: Patient Interactions and Management
A patient in her first trimester is scheduled for a non-urgent brain MRI with and without contrast. What's the most appropriate action?
Reveal answer and explanation for Question 23 (MR-PC-02)
Answer: B
ACR guidance supports clinically indicated non-contrast MRI in pregnancy but advises against routine gadolinium. A gadolinium-based contrast agent should be considered only when the potential for significant clinical benefit outweighs the unknown fetal risk, with the decision documented under institutional policy and discussed with the referring provider and patient. For a non-urgent exam, that may mean deferring the contrast portion until it is clinically necessary.
- A: pregnancy alone doesn't rule out MRI.
- C ignores the guidance.
- D is an improvised dose change made outside the proper decision process.
Review: Safety and Patient Care must-knows · Sources: ACR Gadolinium Pregnancy Screening Statement; ACR Manual on MR Safety
Question 24 · MR-SA-03 · Safety: MRI Screening and Safety
A patient goes into cardiac arrest on the table in Zone IV. What should the MR technologist do first?
Reveal answer and explanation for Question 24 (MR-SA-03)
Answer: C
ACR guidance calls for removing the patient from Zone IV to a predetermined safe location for resuscitation, starting basic life support as the situation allows. Responders and their equipment are a major projectile risk if they rush into the magnet room, so trained MR personnel have to control access during the emergency.
- A risks turning the defibrillator, oxygen tanks, or other equipment into projectiles.
- B: quenching doesn't make the room instantly safe, delays care, and creates its own hazards (helium release and possible oxygen displacement if venting fails). It isn't the response to a code.
- D delays care.
Review: Safety and Patient Care must-knows · Sources: ACR Gadolinium Pregnancy Screening Statement; ACR Manual on MR Safety; APSF Newsletter on MRI emergencies (citing ACR guidance); ACR, What's New in the Manual on MR Safety (March 2026)
Score your set and sort your misses
| Category | Questions | Your correct |
|---|---|---|
| Image Production | 1, 3, 5, 7, 9, 10, 12, 14, 18, 20, 21, 22 | ___ / 12 |
| Procedures | 8, 11, 13, 15, 16, 17, 19 | ___ / 7 |
| Safety | 2, 6, 24 | ___ / 3 |
| Patient Care | 4, 23 | ___ / 2 |
Twenty-four questions is a small sample. Use your misses to pick what to study next, not to predict your result. A percentage here doesn't convert to ARRT's scaled score.
For each miss, write down the topic, the answer you picked, why it looked right, and the principle that actually decides the question. Then label the miss:
- Knowledge gap: you didn't know the fact or concept. Fix: learn it from a source, then write one new example yourself.
- Relationship gap: you knew the definitions but couldn't predict how one change affects another. Fix: drill the tradeoff table.
- Application gap: you knew the concept but picked the wrong action for the situation. Fix: practice with scenario questions, not flashcards.
- Reading error: you missed a word like "first," "except," or a unit. Fix: slow down on the stem, not the options.
Once you can explain why the right answer is right without looking, retire the question. Re-answering a question you've memorized doesn't prove much.
Your 6-week study plan
Assumption: about 8–10 hours a week, roughly 50–60 hours total. That's our editorial estimate, not an ARRT requirement. Adjust it to your baseline and your test date. The order follows ARRT's weights: heaviest category first, then the rest, then mixed timed practice.
| Week | Focus | Do this | You're done when… |
|---|---|---|---|
| 1 | Physical Principles of Image Formation (40 Qs) | Work the practice set and log misses. Study magnetism, the Larmor equation, T1, T2, and T2*, spatial encoding (slice, phase, frequency), k-space, instrumentation, artifacts, and quality control. | You can explain why each common artifact happens and name two fixes |
| 2 | Sequence Parameters and Options (36 Qs) | Drill the tradeoff table until you can fill it in from memory. Work at least 10 scan-time and SNR problems. | You can predict what any single parameter change does to SNR, resolution, scan time, and contrast |
| 3 | Data Acquisition, Processing, and Storage (30 Qs) | Cover pulse sequences (spin echo, fast spin echo, inversion recovery, gradient echo, echo planar, DWI, SWI, perfusion, MRA methods), k-space filling, the Nyquist theorem, post-processing (MIP, MPR, subtraction, ADC maps), and informatics. | You can match each sequence to its clinical purpose and its main weakness |
| 4 | Safety (21 Qs) + Patient Care (16 Qs) | Cover screening and device labels, static/RF/gradient field effects, burns, zones and personnel levels, quench and emergencies, gadolinium agents and reactions, kidney-function screening, ethics and legal topics, and infection control. | You can walk through a code, a quench, and a contrast reaction step by step |
| 5 | Procedures (57 Qs): Neurological (25), then Musculoskeletal (17), then Body (15) | Use the procedure checklist for each exam on ARRT's list. | You can name the planes, coil, landmark, key sequences, and contrast effect for each listed exam |
| 6 | Mixed timed practice + repair | Do timed mixed sets at about one minute per question. Re-drill your two weakest subcategories. Do a final pass through ARRT's content outline to catch topics you never touched. Spend the last two days on formulas and safety. | Your error log shows no repeat misses on the same concept |
Adapting it:
- Four weeks out: combine weeks 1–2, keep week 3, combine weeks 4–5, and keep week 6. Do a few physics retrieval questions every day, because Image Production is half the score.
- Two to three weeks out: spend the first half on Image Production and the second half on Safety, Patient Care, and Procedures. Do short mixed timed sets every other day throughout.
- Ten or more weeks out: give Physical Principles and Sequence Parameters two weeks each, and add a second pass through Procedures. Don't add new resources just to fill time.
- Cross-training from radiography: MRI physics is likely the newest material for you, so start weeks 1–3 early. Ethics and infection control will overlap with what you know; MRI contrast and MR safety won't.
- Working full time: two weekday sessions of about 90 minutes plus one longer weekend block covers the weekly hours.
What's on the exam
ARRT's content specifications (page 1) split the 200 scored questions like this:
| Category | Subcategory | Scored questions |
|---|---|---|
| Patient Care | Patient Interactions and Management | 16 |
| Safety | MRI Screening and Safety | 21 |
| Image Production | Physical Principles of Image Formation | 40 |
| Image Production | Sequence Parameters and Options | 36 |
| Image Production | Data Acquisition, Processing, and Storage | 30 |
| Procedures | Neurological | 25 |
| Procedures | Body | 15 |
| Procedures | Musculoskeletal | 17 |
| Total | 200 |
The exam at a glance
| Scored questions | 200 in both pathways |
|---|---|
| Primary-pathway administration | 230 total items: 200 scored + 30 pilot; 230 minutes test time; 250-minute appointment |
| Postprimary-pathway administration | 220 total items: 200 scored + 20 pilot; 210 minutes test time; 230-minute appointment |
| Pilot questions | Unscored and not identified during the exam |
| Question formats | Mostly multiple choice with one best answer. Some ask you to select multiple answers, sort options into order, select an image, or answer after a short video clip. |
| Passing standard | Scaled score of 75 on a 1–99 scale. That isn't 75% correct. |
| Section scores | Reported from 0.1 to 9.9 as feedback only. They aren't averaged into your total. |
| Attempts | Three attempts within three years of the date your first exam window opens. Each attempt needs a new application and fee. |
| Calculator | Basic and scientific calculators are built into the exam. Staff can give you a four-function handheld on request. You can't bring your own. |
| Where | Pearson test centers |
| Application fee | $225 for the primary pathway or postprimary using an ARRT supporting credential; $450 for postprimary MRI using an NMTCB or ARDMS supporting credential |
Sources: 2026 Primary Eligibility Pathway Handbook (p. 14 attempts; p. 37 formats; p. 40 calculators; p. 43 MRI timing; pp. 53–54 scoring); 2026 Postprimary Eligibility Pathway Handbook (p. 47 MRI timing); ARRT application fees.
Why do the pathway totals differ? ARRT's current 2026 handbooks publish different MRI administration tables. The Primary Handbook lists 30 pilot items and 230 minutes of test time; the Postprimary Handbook lists 20 pilot items and 210 minutes. ARRT's current content specifications still establish the same 200 scored questions and category weights used on this page. Check the handbook for your pathway and your Candidate Status Report when scheduling rather than assuming the other pathway's timing applies to you.
Parameter tradeoffs and a worked example
ARRT's outline says Sequence Parameters questions test how parameters interact and how they affect SNR, contrast-to-noise ratio, spatial resolution, and acquisition time. So study tradeoffs, not isolated definitions. Each row below assumes you change one thing and hold everything else constant.
| If you increase… | SNR | Spatial resolution | Scan time | Also watch |
|---|---|---|---|---|
| TR | Up | No change | Up (proportional in spin echo) | Less T1 weighting; more slices fit per TR |
| TE | Down | No change | No change | More T2 weighting |
| NEX / NSA (doubled) | Up about 1.41× | No change | Doubles | An expensive way to buy SNR |
| Slice thickness | Up | Down (through-plane) | No change | More partial volume averaging |
| FOV (matrix fixed) | Up | Down | No change | Less risk of aliasing |
| Phase matrix (FOV fixed) | Down | Up | Up (proportional) | — |
| Receive bandwidth | Down | No change | No change | Less chemical shift and metal distortion; shorter minimum TE possible |
| Echo train length (FSE) | — | No change | Down (time ÷ ETL) | Contrast set by the effective TE |
| VENC (phase contrast) | — | — | — | Too low causes velocity aliasing; too high loses sensitivity to slow flow |
Sources: Graessner, MAGNETOM Flash (averaging and phase steps); Korosec, ISMRM (FSE time); AuntMinnieEurope artifact review and Hargreaves et al., ISMRM (bandwidth); MRIquestions on VENC.
Worked example: buying resolution
Your baseline knee sequence is fast spin echo with TR 4000 ms, phase matrix 256, NEX 2, and ETL 16. Scan time is (4000 × 256 × 2) ÷ 16 = 128,000 ms, or 2:08.
The radiologist wants sharper images in the phase direction without a longer scan. You raise the phase matrix to 384 and drop NEX to 1:
- Scan time: (4000 × 384 × 1) ÷ 16 = 96,000 ms, or 1:36. Faster than baseline.
- Resolution: smaller pixels in the phase direction, which is what was asked for.
- SNR: down on two counts. Smaller voxels hold less signal, and halving NEX alone cuts SNR to about 0.71× (1 ÷ √2).
If the images now look grainy, put NEX back to 2. Time becomes (4000 × 384 × 2) ÷ 16 = 192,000 ms, or 3:12. That's the lesson: resolution costs you either time or SNR, and exam questions often ask you to name which.
Artifacts: cause, look, and fix
| Artifact | Cause | What you see | Fixes (and the cost) |
|---|---|---|---|
| Aliasing (wrap) | Anatomy outside the FOV in the phase direction | Body parts folded onto the opposite side | Widen the phase FOV or add phase oversampling (longer scan); swap phase and frequency directions |
| Chemical shift | Fat precesses slightly slower than water, so it's mapped a few pixels off along the frequency direction | Bright band on one side of a fat–water border, dark band on the other | Raise the receive bandwidth (lower SNR); fat suppression; swap directions |
| Motion and flow ghosting | Breathing, pulsation, or movement during the scan | Repeated ghost copies or blurring along the phase direction | Breath-holds, gating or triggering, saturation bands, swapping the phase direction |
| Magnetic susceptibility | Metal or air–tissue interfaces distort the local field | Signal voids and warped anatomy | Fast spin echo instead of gradient echo, higher bandwidth, shorter TE |
| Zipper | Outside RF entering through a breached RF shield (often the door) | A noisy line at one fixed position on every image | Close the door; report shield damage |
Sources: RadioGraphics (2022); AuntMinnieEurope; ISMRM 2016 on metal artifacts. ARRT's outline lists more artifacts than this table, including truncation, cross-talk, cross-excitation, moiré, parallel imaging artifacts, eddy currents, and dielectric effect, so learn the cause and fix for each one on that list.
Formulas worth memorizing
- Larmor frequency: f (MHz) = 42.58 × B₀ (T). That's 63.87 MHz at 1.5 T and 127.74 MHz at 3.0 T. (Hornak)
- Fat–water frequency difference: about 3.5 ppm × the Larmor frequency, so roughly 224 Hz at 1.5 T and 447 Hz at 3.0 T. (Reeder, ISMRM)
- Chemical shift in pixels: fat–water difference (Hz) ÷ bandwidth per pixel (Hz/pixel). Example: 224 ÷ 130 ≈ 1.7 pixels.
- 2D spin echo or gradient echo scan time: TR × phase steps × NEX.
- Fast spin echo scan time: (TR × phase steps × NEX) ÷ ETL. (Korosec, ISMRM)
- 3D scan time: TR × phase steps × NEX × slice-encoding steps. (Graessner)
- SNR and averaging: SNR rises with the square root of NEX.
- Inversion time to null a tissue: TI ≈ 0.69 × that tissue's T1, when TR is much longer than T1. For fat at 1.5 T (STIR), that's roughly 150–180 ms. FLAIR uses a much longer TI to null CSF. (MRIquestions)
Convert milliseconds to seconds before minutes, and sanity-check every answer. A fast spin echo sequence runs minutes, not half an hour.
Safety and Patient Care must-knows
Together these are 37 scored questions, and they're where real-world mistakes hurt people. Know these cold:
- Zones: Zone I is public. Zone II is where screening happens. Zone III is restricted and includes the control area; Level 2 MR personnel control access to it. Zone IV is the magnet room, always inside Zone III.
- Personnel: Level 1 MR personnel have basic safety training. Level 2 personnel have extended training, including RF burns and peripheral nerve stimulation, and handle screening before Zones III and IV.
- Device labels: MR Safe (no known hazards in any MR environment), MR Conditional (no known hazards only under stated conditions), MR Unsafe. Passive metallic implants can't be MR Safe. Conditional labels apply at the field strengths tested; a device cleared for 3 T can't be assumed safe at 1.5 T.
- Burns: most MRI injuries are thermal. Change patients into MR-safe gowns, pad skin-to-skin contact points and the bore, keep cables off the skin, and never let cables form loops. Remove unused coils and leads from the bore; never scan with unconnected coils left in place.
- Field effects: static field (pulling and twisting forces, projectiles); RF field (heating measured as SAR, burns); gradient field (loud noise, peripheral nerve stimulation). Use hearing protection; ACR recommends earplugs plus muffs whenever possible.
- Quench: rapid boil-off of the liquid helium that cools the magnet. If venting fails, helium can displace oxygen in the room.
- Emergencies: for a code, move the patient out of Zone IV to a predetermined safe area, starting basic life support as the situation allows. Don't let unscreened responders or equipment in.
- Contrast: know ACR group II agents, eGFR screening, reaction recognition and response, and that gadolinium isn't routine in pregnancy.
- Communication: time-outs and communicating critical findings to the care team were added to ARRT's Patient Care outline in 2025.
Sources: ACR Manual on MR Safety and its March 2026 summary of changes; Shah and Aran, Cureus (2023); burn-prevention table, J Med Radiat Sci; ARRT 2024 notice of MRI document changes. Your facility's policy and MR safety leadership govern real cases.
Procedures: what to prioritize
Neurological is the largest Procedures subcategory (25 questions), then Musculoskeletal (17), then Body (15). ARRT's content specifications (pp. 8–9) list each exam that can appear, from brain, spine, and vascular head and neck through liver, MRCP, pelvis, and every major joint.
For each exam on that list, you should be able to answer:
- Anatomy and planes: Which planes show the anatomy best, and what are you looking for?
- Pathology and protocol: What condition is suspected, and which sequences answer the question?
- Patient considerations: What changes for a child, an older adult, a larger patient, or a trauma patient?
- Setup: Which coil, which orientation, and where do you landmark? Does it need cardiac or respiratory gating?
- Contrast: If contrast is given, how does it change the images?
Work through neurological exams first, especially brain, spine, and vascular head and neck. Learn anatomy and protocol together; an isolated anatomy label rarely answers an exam question by itself.
What's changing in 2028
On July 17, 2026, ARRT proposed new MRI documents with a scheduled implementation date of February 1, 2028. Comments closed October 2, 2026. These are still proposals: ARRT's Board of Trustees has to review them, and the committee may revise them.
The biggest proposed changes:
- Safety would grow from 21 to 40 scored questions, and you'd need a passing score on the Safety section as well as on the whole exam.
- Spectroscopy would be removed from Image Production, and brain spectroscopy from Procedures.
- CINE would move under Pulse Sequences, and image quality would move into Sequence Parameters and Options.
- New procedures would be added: brain for known tumor or metastases, CSF leaks, spinal trauma, iron quantification, and scapula.
If you test before February 1, 2028, the current specifications on this page apply. Don't study the proposed weights as if they're in effect now. If you're retaking near that date, check which version covers your exam window.
Which pathway are you on?
Both pathways use the same current MRI content specifications and 200 scored-question blueprint. What differs is how you qualify, and ARRT's current 2026 handbooks also list different pilot-item counts and test times by pathway.
Primary pathway (MRI program graduates):
- Complete an ARRT-verified MRI educational program, including its didactic and clinical competency requirements.
- Earn an associate degree or higher.
- Apply within three years of finishing the program; you can apply up to 90 days before you finish.
Postprimary pathway (already credentialed):
- Hold a supporting credential: an ARRT credential in a supporting discipline such as Radiography, or an NMTCB nuclear medicine or ARDMS sonography credential.
- Complete at least 16 hours of structured education, with at least one credit in each of the four content categories, within the 24 months before you apply.
- Document at least 125 repetitions across at least 21 different procedures (3–6 repetitions each), verified by an R.T. or an interpreting physician.
- Document training in eight MRI safety areas and competence in seven quality control activities.
- Remote scanning doesn't count; you must be physically present for each procedure.
Sources: 2026 Primary Handbook (pp. 6–7, 22–23); 2026 Postprimary Handbook (pp. 5–8); MRI Clinical Experience Requirements; ARRT MRI credential page.
Not the same as I.A.(MR). ARRT's newer Imaging Assistant, Magnetic Resonance credential is a separate exam for assistants who support MRI technologists with safety screening, patient care, and positioning. They don't scan. It isn't the MRI registry.
Frequently asked questions
Is the postprimary MRI exam different from the primary one? The scored content is the same: both pathways use the current MRI content specifications and 200 scored-question blueprint. ARRT's current 2026 handbooks do, however, list different pilot-item counts and test times by pathway, explained above.
Does a scaled 75 mean I need 75% correct? No. ARRT converts the number you answer correctly to a 1–99 scale that adjusts for differences in difficulty between exam versions. ARRT doesn't publish a fixed percentage, so be skeptical of any "you need X% correct" claim.
Can I tell which questions are pilots? No. They're mixed in and look like every other question, so treat every question as scored.
Does ARRT recommend any review books or courses? No. ARRT doesn't endorse review programs, mock registries, or study guides, and it warns against materials that claim to include real ARRT questions. Its content specifications are the official topic list (2026 Primary Handbook, p. 38).
How can I tell if a prep resource is current? Check that it uses the specifications effective February 1, 2025; shows the four categories with 16, 21, 106, and 57 scored questions; keeps the proposed 2028 changes separate from the current exam; and doesn't treat a practice percentage as an ARRT scaled score. A resource that presents 40 Safety questions as the current exam is out of step.
When do I get my results? You see a preliminary scaled score on screen at the test center, but it isn't final. ARRT says that within about 10 business days you'll receive an email telling you that your official score report is available in your online ARRT account. More: ARRT exam results.
What if I don't pass? You can reapply within your three-attempt, three-year eligibility period. Compare the current specifications against the ones you studied before you start again. More: ARRT exam retake policy.
Keep going
- Ready to book? How ARRT exam registration works
- After you pass: ARRT continuing education requirements
- Official ARRT documents: MRI credential page · MRI content specifications · Handbooks · Application fees
Sources and verification
Last verified October 27, 2026. The current exam blueprint, pathway-specific administration tables, eligibility requirements, fees, and 2028 proposal status on this page were checked against these ARRT documents and pages:
- MRI Examination Content Specifications, Board approved January 2024, effective February 1, 2025
- 2026 Primary Eligibility Pathway Handbook and 2026 Postprimary Eligibility Pathway Handbook
- MRI Structured Education Requirements and MRI Clinical Experience Requirements
- Application fees, the 2024 notice of MRI document changes, and the July 17, 2026 notice of proposed changes
- ARRT exam scoring for preliminary and official-score timing
- ACR Manual on MR Safety, ACR Manual on Contrast Media, and the ACR Gadolinium Pregnancy Screening Statement for the safety and contrast items that cite ACR guidance
The teaching behind each practice question is linked inside its explanation. Those sources support the physics, safety, and clinical principles; ARRT's specifications only show which topics are tested. ARRT hasn't reviewed this page, and no MRI technologist or radiologist has reviewed the practice questions. Fees, handbooks, and specifications change, so confirm with ARRT before you apply.
Independence and trademarks: Castleport Test Prep is an independent exam prep publisher, not affiliated with, endorsed by, or approved by The American Registry of Radiologic Technologists (ARRT). Exam and credential names are used to identify their subjects, and trademarks belong to their respective owners. All practice questions are original and unofficial. This page is educational; it isn't medical advice and doesn't replace your facility's policies or your MR safety leadership.