ARRT Radiography Exam Pass Rates by Program: 2026 Data

By Castleport Test Prep Editorial Team Published: August 8, 2026 · Last verified: August 8, 2026

Castleport Test Prep Research is an independent research and reference resource.

The finding, up front

ARRT radiography exam pass rates by program are not one single number. The public record has three different measures, and they do not mean the same thing.

ARRT’s 2025 report shows an 85.8% first-time radiography pass rate across 14,070 candidates. JRCERT’s latest national report shows an 89% five-year credentialing pass rate for radiography programs reporting on the 2024 graduating class.

The two national numbers are 3.2 percentage points apart because they count different people over different spans of time. JRCERT’s latest fixed count also shows 611 accredited radiography programs as of December 2025. That is a dated snapshot, not a promise that the live directory still has exactly 611 today.

Putting three JRCERT reports side by side shows something else: its national five-year credentialing pass rate moved from 93% to 91% to 89%, while completion and job placement moved up.

Table 1 — JRCERT national radiography outcomes, three graduating classes

The reported completion rate rose. The reported job-placement rate rose. The reported credentialing pass rate fell.

Each JRCERT reading is two points lower than the one before it. JRCERT does not give a reason, so this page does not invent one. The direction also differs from ARRT’s national first-time candidate rate, which rose 1.4 points across its 2023, 2024, and 2025 report editions.

ARRT gives its own warning:

“Data in this report are not intended to be used as the sole source for educational program evaluation.”

Source: ARRT Annual Exam Report – 2025, page 1.

One more thing before you scroll. A common math mistake can turn a published 88.2% five-year rate into 36.0%. That is a 52.2-point error. The live example is in “Four ways these tables fool people.”

ARRT radiography exam pass rates by program: the three official measures

Answer capsule: ARRT publishes a national first-time candidate pass rate and an anonymous distribution of program mean scaled scores. JRCERT publishes a named five-year credentialing pass rate for each accredited program. A program mean score is not a pass rate, and ARRT’s public annual reports do not name schools.

First, the two organizations:

ARRT is the American Registry of Radiologic Technologists. It writes and scores the exam. It certifies and registers people.

JRCERT is the Joint Review Committee on Education in Radiologic Technology. It accredits programs. JRCERT says it is the only agency recognized by the U.S. Department of Education for accrediting traditional and distance-delivery programs in radiography, radiation therapy, magnetic resonance, and medical dosimetry.

Now the three measures.

Table 2 — Three official measures that are often called a pass rate

For one named program, use the third row. It is the only measure that attaches a credentialing percentage to a school.

Blur these measures and a scaled score can quietly turn into a percentage. A “95% credentialing rate” and a “program mean score of 89” cannot be compared as if they were the same thing.

What is the national ARRT radiography pass rate right now?

Answer capsule: As of August 8, 2026, ARRT’s latest Annual Exam Report is the 2025 edition. It gives a published one-decimal first-time radiography pass rate of 85.8% across 14,070 first-time candidates. Across the last three report editions, the rate rose from 84.4% to 85.8%, while first-time candidate volume grew 22.5%.

Table 3 — ARRT radiography exam results, 2023 to 2025 report editions

The four exam-volume categories add to the printed total in all three editions. That confirms the table’s internal arithmetic. It does not prove every outside interpretation of the report.

First-time candidate volume rose by 2,581 people, from 11,489 to 14,070. That is 22.5% growth across the three report editions.

Two warnings before quoting an ARRT number

Warning 1: the 2025 report prints 86% and 85.8%. The summary table rounds the radiography pass rate to 86%. The radiography section gives 85.8%. Use 85.8% when you want the more precise published figure, and name the report edition.

Warning 2: the report introductions have a year-label problem. The 2023, 2024, and 2025 editions all say that their specific results are for 2024, even though their data are different. The first-time candidate counts alone are 11,489, 12,665, and 14,070. Cite the report edition, not the repeated sentence in the introduction.

How do program mean scores compare across the country?

Answer capsule: ARRT publishes an anonymous distribution of each program’s mean scaled score. In the 2025 report, 717 radiography programs were represented. The median program mean was 84. The 5th-percentile mean was 75, and the 95th-percentile mean was 89. These are scaled scores, not pass rates.

Table 4 — Radiography program mean scaled-score percentiles

A scaled score is not a percent correct. ARRT uses a 1-to-99 scale and sets 75 as the passing score. The annual reports say radiography required about 67% correct to reach the passing standard. The exact raw score can differ by exam form because ARRT adjusts for differences in form difficulty.

Four things the table shows

1. The middle half of program means fits inside five points.

The 25th-percentile program mean was 81. The 75th-percentile mean was 86. That five-point band holds the middle 50% of the 717 program means represented in the 2025 report.

2. Individual scores spread more widely than program means.

Table 5 — Width of the 2025 ARRT score distributions

This is a statement about the width of two published distributions. It does not prove that a school has no effect, and it does not explain why individual scores vary.

3. All 19 published program percentile points meet or exceed 75.

The 5th-percentile program mean is exactly 75, the passing scaled score. That does not mean every candidate at those programs passed. A program mean is not a program pass rate.

4. The program distribution changed very little from 2024 to 2025.

Seventeen of the 19 published percentile points are identical. Only the median moved from 83 to 84, and the 5th percentile moved from 74 to 75.

Where do I find one specific program’s pass rate?

Answer capsule: Use JRCERT’s Find a Program directory, open the program’s detail page, and scroll to the bottom. Then check the program’s own website. JRCERT requires annual public posting and requires the sample size to be shown with each measure.

There are two places. Compare both. Annual update timing can make one page newer than the other.

Place one: JRCERT’s Find a Program directory. The search results show program type, name, city, state, and award type. Pass rate is not a results column. Open the individual program record and look at the Program Effectiveness Data at the bottom.

Place two: the program’s own website. JRCERT says the data must be on the program’s main page or clearly linked from that page, cannot be buried through many layers, and must be updated at least once a year. JRCERT reviews program sites and requires correction plus a progress report when it finds gross differences from the annual report.

What to copy from a program page

1. The five years covered
2. The number who passed on the first attempt
3. The number who tested within six months
4. The stated five-year rate
5. Whether the page says ARRT only or ARRT or an unrestricted state exam
6. The date you checked

Table 6 — Verified named-program examples

These are clear examples, not a best-to-worst list.

The arithmetic checks:

• 94 ÷ 99 = 94.95%, shown as 95% on the Sauk Valley page.
• 102 ÷ 124 = 82.26%, shown as 82.3% on the Charles R. Drew page.
• 50 ÷ 58 = 86.21%. The IU Kokomo page displays 86%; this table reports the count-based result to one decimal place.
• 118 ÷ 120 = 98.33%, shown as 98.33% on the IU Northwest page.

That small IU Kokomo formatting difference is harmless because the counts are visible. The counts are the part you can test.

What does JRCERT’s credentialing pass rate count?

Answer capsule: JRCERT counts graduates who passed a qualifying exam on the first attempt, divided by graduates who took that exam within six months of graduation, across the most recent five consecutive years. The benchmark is 75%. The qualifying exam can be the ARRT certification exam or an unrestricted state licensing exam.

The rule appears in Standard Six, Objective 6.1 of JRCERT’s 2021 Radiography Standards. A program must maintain a five-year average credentialing pass rate of at least 75% at first attempt within six months of graduation.

Table 7 — What is inside and outside the JRCERT calculation

The correction most pages miss

The search phrase says “ARRT pass rate.” But JRCERT’s official definition allows either the ARRT exam or an unrestricted state licensing exam. A JRCERT five-year credentialing rate is not automatically an ARRT-only rate.

Some programs narrow the wording and say ARRT. Charles R. Drew does. Others print JRCERT’s broader definition. Read the line above the table before calling a rate ARRT-only.

One rule that makes the math checkable

JRCERT requires program effectiveness data to identify the sample size for each measure. That means the numerator and denominator should be visible, not just the percentage.

That one rule lets a reader redo the math. It also exposes the zero-denominator trap described below.

Why do the ARRT and JRCERT numbers differ?

Answer capsule: ARRT’s 85.8% and JRCERT’s 89% measure different groups over different time windows. Neither number cancels the other. ARRT reports first-time candidates in one report edition. JRCERT reports a five-year program measure limited to graduates who tested within six months.

Table 8 — ARRT and JRCERT, side by side

The two different 75s

ARRT’s 75 is a scaled score. It is the passing line for an individual exam. It is not 75% correct.

JRCERT’s 75 is a percentage. It is the minimum five-year credentialing pass rate a program must maintain.

Same digits. Different measures.

Do not subtract 611 from 717

ARRT’s 2025 report says 717 educational programs were represented in its program-mean table. JRCERT’s fixed December 2025 snapshot counted 611 accredited radiography programs.

Do not subtract the two and label the difference. The counts come from different systems, use different inclusion rules, and describe different moments. Publish each number with its definition and leave the subtraction alone.

Four ways these tables fool people

Answer capsule: The biggest errors are averaging yearly percentages instead of pooling the counts, treating every displayed zero as failure, trusting the “Year 1” position, and reading a five-year rate as if it were the latest class. The first mistake can change the answer by more than 50 points.

Every trap below comes from a current official or institutional page.

Trap 1: averaging the five percentages

JRCERT’s rate is built from pooled counts: add all first-attempt passes, add all qualifying test takers, then divide once.

Do not take a simple average of the five displayed yearly percentages.

Indiana University Northwest’s radiation therapy page gives the cleanest live example. It uses the same JRCERT definition and table structure as radiography.

Table 9 — The live 52.2-point averaging error

A simple average of the five displayed percentages is:

(0 + 100 + 0 + 80 + 0) ÷ 5 = 36.0%

The pooled rate is:

15 ÷ 17 = 88.24%, shown as 88.2%

The gap is 52.2 percentage points.

Three rows are 0 of 0. No one tested in those rows, so they add no people to the pooled denominator. The program’s pooled total is mathematically correct. The printed 0% beside 0 of 0 is what makes a simple average fail.

The rule: pool the counts. If a page gives only percentages, the five-year math cannot be checked.

Trap 2: a displayed zero is a flag, not a conclusion

JRCERT says a newly accredited program may appear with a pass rate of “0” when it does not yet have enough history. A yearly row can also show 0 of 0 when no graduate tested inside the window.

But not every zero is harmless:

• 0 of 0 means there is no result for that row.
• A new-program summary marked 0 may mean the five-year history does not exist yet.
• 0 of a positive number is a real 0% for that denominator.

Read the counts before writing what the zero means.

Trap 3: “Year 1” can point in opposite directions

Indiana University Kokomo and Indiana University Northwest label 2024 as Year 1, with the newest year first. JRCERT’s current 2022–2026 blank form labels 2022 as Program Year 1, with the oldest year first.

Same field name. Opposite direction.

Source: IU Kokomo program data, IU Northwest program data, and JRCERT’s 2022–2026 Program Effectiveness Data form. Verified August 8, 2026.

Read the printed calendar year. Never trust the row position by itself.

Trap 4: a five-year average can hide one weak class

Table 10 — Current radiography examples with the yearly spread exposed

IU Kokomo’s five-year rate is above the 75% floor, yet one class was at 64%. Charles R. Drew’s current five-year rate is 82.3%, yet its 2025 class was at 62%.

The five-year figure is accurate. It is also slow. Read the individual years.

How much can one weak year move a five-year average?

Answer capsule: Under an equal-class-size example, four years at 100% plus one year at 0% still produces an 80% five-year rate. Two 0% years produce 60%. Real JRCERT rates pool student counts, so unequal class sizes can change the result.

The table below is an illustration, not a claim about any named program. It assumes the same number of test takers each year.

Table 11 — Five-year smoothing under equal cohort sizes

This cuts both ways. A sudden weak year may take time to pull the five-year number below the floor. A real improvement may also take time to pull the number back up.

The published rate describes the most recent five-year reporting window in one pooled number. It does not tell you, by itself, whether a recent staffing, clinical-site, admissions, or curriculum change has helped. The annual rows show change sooner.

How many JRCERT-accredited radiography programs are there, and where?

Answer capsule: JRCERT’s latest fixed national snapshot counted 611 accredited radiography programs as of December 2025. Associate-degree programs made up 457, or 74.8%. Florida and Texas had 39 each. Vermont was printed at zero. Alaska was omitted from the table and is coded as zero in the downloadable file with that inference clearly marked.

Table 12 — All 611 programs by sponsor and terminal award

The largest single cell is the community-college associate degree: 270 programs, or 44.2% of all 611.

Hospital programs look different. 85 of 98 hospital-run programs, or 86.7%, award a certificate. Those 85 hospital certificate programs make up 85% of all certificate programs in the table.

JRCERT also reports 11,832 clinical settings tied to the 611 radiography programs. That is about 19.4 clinical settings per program, though the real number varies by school.

Where the programs are concentrated

The ten states below hold 297 programs, or 48.6% of the December 2025 total.

Table 13 — Ten states with the most JRCERT-accredited radiography programs

Among the 50 states, the median was 10 programs. Twelve states or jurisdictions had two or fewer once the District of Columbia, Puerto Rico, and the Alaska inference are included.

Vermont is printed as zero. Alaska is not printed at all. The source table lists every other state plus the District of Columbia and Puerto Rico, and its total reconciles without Alaska. The downloadable state file therefore adds Alaska as zero but labels the row inferred from omission, not directly printed.

The reported pipeline grew faster than the program count

Table 14 — JRCERT radiography program trend, 2022 to 2024

The number of reporting programs rose 1.5% while reported graduates rose 17.3%. Most of the net growth came from larger reported cohorts, not a large increase in the number of reporting programs.

That points in the same direction as ARRT’s data, where first-time candidate volume rose 22.5% across the 2023 to 2025 report editions.

How should I compare two radiography programs fairly?

Answer capsule: Match the five-year window, confirm the exam scope, and use the raw counts. A 100% rate based on four test takers carries less information than a rate based on 120.

Use these five checks in order:

1. Same five years? Two pages can both say “five-year average” while covering different classes.
2. Same exam scope? One page may be ARRT-only. Another may use the broader ARRT-or-state definition.
3. First attempt only? JRCERT’s measure should be. Confirm the wording.
4. Are the counts shown? A percentage without a numerator and denominator cannot be checked or pooled.
5. How large are the classes? One student changes a ten-person class by ten points. One student changes a 100-person class by one point.

Why pass rate should not be the only test

ARRT says its data should not be the sole source for program evaluation. Its report points to class size, candidate characteristics, admissions criteria, and curriculum emphasis as factors that can affect program-level results.

The 2025 distribution adds one more reason for care. Individual candidate scores covered a much wider range than program mean scores. That is a descriptive finding, not proof of cause.

Use the credentialing rate. Then also check completion rate, job placement rate, current accreditation status, award length, class size, cost, and whether the clinical sites are places you can reach on the required schedule.

Can I sort radiography programs from best to worst?

Answer capsule: Not fairly from these public numbers alone. The reporting windows can differ. The exam scope can differ. Small classes swing hard. ARRT’s public program table gives anonymous mean scores, while JRCERT’s named figures are pooled five-year credentialing rates.

A simple high-to-low list would hide four problems:

• Program windows do not always match.
• Some pages use ARRT-only wording; others use ARRT-or-state wording.
• Small denominators make percentages jump.
• Half of the 2025 ARRT program means sit inside a five-point band, so many programs are close together on that measure.

A sorted list would make the data look more precise than it is. The honest use is to compare matched windows, matched definitions, raw counts, and the other outcomes that matter to a student.

How we built this research

Answer capsule: Castleport Test Prep Research reviewed three ARRT Annual Exam Reports, JRCERT standards and public-data rules, three JRCERT outcome snapshots, the December 2025 program landscape, the 2022–2024 trend table, the current JRCERT form, and current institutional program pages. Every figure with printed counts was recalculated.

What we did

1. Read ARRT’s Annual Exam Reports for 2023, 2024, and 2025.
2. Recorded radiography candidate volume, mean scaled score, published one-decimal pass rate, approximate raw percent needed to pass, and programs represented.
3. Transcribed all 19 radiography program mean percentile points from each edition.
4. Checked that the four exam-volume categories add to the printed total in every edition.
5. Read JRCERT’s Program Effectiveness Data definitions, Program Effectiveness FAQs, 2021 Radiography Standards, current Find a Program page, and current 2022–2026 data form.
6. Pulled national outcome figures for graduating classes 2015, 2020, and 2024 from three JRCERT Organizational Reports.
7. Pulled December 2025 state counts, sponsor-and-award counts, clinical settings, and the 2022–2024 trend table.
8. Reconciled sponsor totals and state totals to JRCERT’s printed total of 611.
9. Opened current program pages and recalculated every rate that had a printed numerator and denominator.
10. Kept candidate pass rate, program mean scaled score, and JRCERT five-year credentialing rate in separate fields from start to finish.

Calculation rules

• Percent change: (new − old) ÷ old
• Percentage-point change: new percentage − old percentage
• Program credentialing rate: pooled passes ÷ pooled test takers, never the simple mean of yearly percentages
• Share of programs: category count ÷ 611 for the December 2025 snapshot
• Blank values: never converted to zero
• ARRT-only label: used only when the source page says ARRT
• Rounding: count-based calculations are shown to one decimal place unless the source gives more precision
• Printed aggregate without counts: transcribed as published, not presented as independently re-derived

Source order

1. ARRT official reports and scoring pages
2. JRCERT official reports, standards, forms, directory, and public-data rules
3. Official institutional program pages
4. No third-party page was used as factual evidence

Limitations: what the data does and does not show

We would rather you see these limits here than find them later.

1. ARRT’s public annual reports do not name schools. The program mean table is anonymous.
2. ARRT’s program percentile is a mean scaled score, not a pass rate. Do not turn a mean score of 89 into an 89% pass rate.
3. JRCERT’s program data is self-reported. JRCERT says programs submit it through the annual report, and the current Organizational Report says staff reviewed submissions for accuracy and followed up on concerns.
4. A JRCERT credentialing rate may include an unrestricted state licensing exam. It is not always ARRT-only.
5. The national measures cover different spans. ARRT’s is tied to one report edition. JRCERT’s is a rolling five-year program measure.
6. The JRCERT readings are rolling five-year measures. They are useful snapshots, not three independent single-year samples.
7. The data does not explain cause. This page does not claim to know why one measure moved up while another moved down.
8. Small cohorts make rates unstable. Always read the counts.
9. A displayed zero needs its denominator. 0 of 0 is no result; 0 of a positive number is a real zero rate.
10. ARRT’s 717 and JRCERT’s 611 are different universes. Do not subtract them and name the remainder.
11. The December 2025 program count is a snapshot. The live JRCERT directory can change after that date.
12. The named examples are selected for clear, current data. They are not a random sample of all programs.
13. Program pages update. Recheck any named figure at its source before using it for a current decision.
14. ARRT’s report introductions contain a repeated year label that conflicts with the changing data. Cite the report edition.
15. JRCERT calls its 2022–2024 trend figures approximate. It also warns that timing can create small differences from other tables.
16. Alaska’s zero is inferred from omission. Vermont’s zero is printed. The downloadable file marks that difference.

Frequently asked questions

Does ARRT publish pass rates by school?

No. In the 2023, 2024, and 2025 Annual Exam Reports reviewed here, ARRT publishes a national first-time pass rate, state-level results, and an anonymous distribution of program mean scaled scores. Named program credentialing rates come from JRCERT or the program’s own site.

What is the ARRT radiography pass rate?

The published one-decimal first-time radiography pass rate is 85.8% across 14,070 first-time candidates in ARRT’s 2025 Annual Exam Report. The report’s summary table rounds that figure to 86%.

What is a good credentialing pass rate for a radiography program?

JRCERT’s minimum is a 75% five-year average. Its latest national radiography figure is 89% for programs reporting on the 2024 graduating class. Compare the program’s counts, years, and exam wording before deciding what its number means.

Is a 75 on the ARRT exam the same as 75% correct?

No. ARRT uses a scaled score from 1 to 99, and 75 is the passing score. ARRT says a scaled 75 is not 75% correct. Its 2025 radiography summary gives about 67% correct as the approximate level needed to pass.

Why does a program show a 0% pass rate?

Check the counts. A row showing 0 of 0 has no test result. JRCERT also says a new program may show a summary zero before it has enough history. But 0 passes out of a positive number of test takers is a real 0% for that group.

How do I find my school’s pass rate?

Search JRCERT’s Find a Program directory, open the program record, and scroll to the Program Effectiveness Data. Then compare it with the program’s own website. Copy the years, passes, test takers, stated rate, and exam definition.

Can I average the five yearly percentages?

No. Pool the counts. On the live Indiana University Northwest radiation therapy table reviewed here, the simple average of the five displayed percentages is 36.0%, while the correct pooled rate is 88.2%. The error is 52.2 points.

How many JRCERT-accredited radiography programs are there?

JRCERT’s latest fixed snapshot counted 611 as of December 2025. Of those, 457 were associate-degree programs, 100 certificate programs, 53 bachelor’s programs, and one master’s program. Use the live directory for a current school search.

Why is JRCERT’s national number higher than ARRT’s?

They count different groups over different time spans. JRCERT limits its measure to graduates of accredited programs who test within six months and pools five years. ARRT reports first-time candidates in one report edition. The 3.2-point difference is not a contradiction.

Is the radiography pass rate going up or down?

Both answers appear in the official data because the measures differ. ARRT’s first-time candidate rate rose from 84.4% to 85.8% across its 2023 to 2025 report editions. JRCERT’s five-year national program figure moved from 93% for the 2015 graduating class to 91% for 2020 and 89% for 2024.

Data files

The core data behind this page is available as plain CSV so the main totals and calculations can be checked:

• ARRT radiography national series, 2023–2025 report editions (CSV)
• ARRT radiography program mean percentiles, 2023–2025 (CSV)
• JRCERT national radiography outcomes, graduating classes 2015, 2020, and 2024 (CSV)
• JRCERT radiography programs by state and jurisdiction, December 2025 (CSV)
• JRCERT programs by sponsor and terminal award, December 2025 (CSV)
• JRCERT radiography program trend, 2022–2024 (CSV)
• ARRT-to-JRCERT metric crosswalk (CSV)
• Verified named-program examples with yearly counts (CSV)

The state file has 52 rows: 50 states, the District of Columbia, and Puerto Rico. JRCERT’s printed table has 51 rows because Alaska is omitted. The Alaska row is added as zero and marked as an inference.

How to cite this page

Citation information:

Castleport Test Prep Editorial Team. “ARRT Radiography Exam Pass Rates by Program: 2026 Data.” Castleport Test Prep Research. Published and last verified August 8, 2026. https://castleporttestprep.com/research/arrt-radiography-exam-pass-rates-by-program/

Primary sources