- What "Passing" Means for the CRR
- The Numbers Behind the Score
- Automatic-Failure Conditions
- How the Test Is Delivered
- The Post-Dictation Timers
- What Historical Pass Rates Tell You
- Turning the Score Into a Training Target
- A CRR-Specific Preparation Sequence
- Eligibility, Fees and Windows
- After You Pass: Maintaining the Credential
- Frequently Asked Questions
- The CRR is one five-minute testimony/Q&A dictation at 200 words per minute, scored for 96% accuracy.
- The transcript must be unedited realtime output; the grading guide permits 40 errors.
- Two drops of 10 or more consecutive words, or a file with no printable data, fail automatically.
- Recent pass rates were 9% (2023), 8% (2024) and 11% (2025) across completed tests submitted for scoring.
What "Passing" Means for the CRR
If you are searching for the CRR passing score, the first thing to understand is that the Certified Realtime Reporter credential from the National Court Reporters Association (NCRA) is not scored like a written exam. There is no scaled score, no 200-to-800 range, no cut score applied to a question bank. There is one performance component: a realtime testimony/Q&A dictation, and your result is determined by how accurately your steno machine and computer-aided transcription (CAT) software render that dictation without any editing.
That makes the "passing score" a transcript accuracy threshold: 96% accuracy on a five-minute dictation read at 200 words per minute. Everything else in the process, from the proctoring setup to the upload timers, exists to make sure that what gets graded is genuinely your unedited realtime output.
If you are still orienting yourself to the credential itself, the explainer pages on what CRR certification is and what CRR stands for cover the basics. This article focuses squarely on the scoring standard.
The Numbers Behind the Score
The standard has four moving parts you should be able to recite from memory: the speed, the duration, the accuracy threshold and the error allowance. Here they are in one place.
| Element | CRR Requirement |
|---|---|
| Content | Testimony/Q&A dictation |
| Dictation length | Five minutes |
| Speed | 200 words per minute |
| Accuracy required | 96%, without editing the realtime transcript |
| Error allowance (per current linked grading guide) | 40 errors |
| Written knowledge exam | None; the skills test is the sole component |
Do the arithmetic and the reality of the standard becomes clear. Five minutes at 200 words per minute is roughly 1,000 words of dictation. A 96% accuracy requirement means the error budget is about 4% of that text, which is consistent with the 40-error allowance in the grading guide. Forty errors sounds generous until you remember that a single untranslated stroke, a conflict, a missing word or a misstroke each counts, and that realtime errors tend to cluster when a speaker accelerates or a string of unfamiliar terms arrives.
Automatic-Failure Conditions
Reaching 96% accuracy on paper does not help you if you trip one of the automatic-failure conditions. According to the grading guide, two situations end the attempt regardless of how clean the rest of the transcript is:
- Two drops of 10 or more consecutive words. A single large drop is damaging; two of them are disqualifying. This is the failure mode of a reporter who loses the thread during a fast passage and stops writing.
- An uploaded file without printable data. If the file you submit does not contain printable content, it cannot be graded. This is a procedural failure rather than a skill failure, which is why practicing the upload sequence matters.
The first condition shapes how you should train. A realtime reporter who panics after a mistake and stops stroking can turn one rough sentence into a 10-word hole. Training should deliberately include recovery: when you hit an unfamiliar term, you write it phonetically and keep moving rather than freezing.
Key Takeaway
Treat "never drop 10 consecutive words" as a separate skill from raw accuracy. Run drills where you intentionally misstroke early, then practice continuing to write cleanly instead of stopping to fix the error.
How the Test Is Delivered
The CRR is taken remotely, and the delivery environment is part of the pass/fail picture because equipment problems can derail an otherwise qualified candidate.
- Platform: RealtimeCoach, with recorded remote proctoring through ProctorU/Record+ using the Guardian browser.
- What you supply: compatible stenographic equipment, CAT software, a webcam, a microphone and wired headphones. Bluetooth headphones are not permitted.
- Exam preview: NCRA offers a preview that checks your equipment and walks you through the proctored process. It is not speedbuilding practice and does not produce an official score.
The preview is easy to underestimate. It will not tell you whether you can hit 96%, but it can expose a wired-headphone problem, a CAT configuration issue or a Guardian browser conflict before the stakes are real. Use it to eliminate every avoidable variable.
The Single Assessed Domain: Realtime Testimony/Q&A
The CRR assesses one skill: producing accurate realtime transcript of testimony and question-and-answer material at speed. It is not a weighted blueprint of topics.
- Q&A format, with alternating questioner and witness speech
- Sustained speed of 200 words per minute for five minutes
- Clean translation straight from your steno output, with no cleanup
- Consistent personal dictionary and briefs that translate without conflicts
For a closer look at how this single domain is framed, see the CRR exam domains guide.
The Post-Dictation Timers
The scoring standard applies to your transcript, but you must get that transcript into the system correctly. After the dictation ends, separate timers govern the submission steps:
- Three minutes to attach your notes.
- Five minutes to attach the unedited transcript.
- Two minutes to paste the final transcript.
Setup and warm-up time are additional to these figures, and the five minutes is the assessed dictation time, not the total length of your appointment. Critically, upload time is not a separate transcription period and is not permission to edit. The transcript you attach must be the unedited realtime output. Practicing the full export-and-attach sequence under a timer is as important as practicing the dictation itself, because a fumbled upload can lose an otherwise passing attempt.
What Historical Pass Rates Tell You
NCRA's published statistics give a sobering picture of how demanding the standard is in practice:
| Year | Pass Rate | Tests Submitted for Scoring |
|---|---|---|
| 2023 | 9% | 345 |
| 2024 | 8% | 281 |
| 2025 | 11% | 304 |
Read these carefully. They describe completed tests submitted for scoring, not first-time-candidate success rates, and they are not a prediction for any individual. A candidate who has spent months in structured realtime training is in a very different position from one attempting the test cold. Still, the pattern is consistent: the 96% unedited standard is a high bar that most submitted attempts do not clear. For a deeper treatment of the data, see the CRR pass rate breakdown and the CRR difficulty guide.
Turning the Score Into a Training Target
The most useful way to think about 96% is as a diagnostic framework. After every practice take, categorize your errors rather than just counting them. The categories that matter for realtime work include:
- Untranslates: strokes your dictionary does not recognize. These point to dictionary gaps, usually names, technical vocabulary or unusual phrasing.
- Conflicts: strokes that translate into the wrong word because multiple entries share an outline. These point to briefs that need disambiguating.
- Misstrokes: hand or timing errors, the kind that grow worse as speed climbs.
- Drops: missing words, especially dangerous because consecutive drops feed the automatic-failure rule.
- Boundary and formatting problems: word-boundary errors, punctuation and speaker-identification issues in Q&A text.
Preparation should rely on original timed dictation and unedited realtime output rather than recalled NCRA test content or multiple-choice drilling, neither of which applies to this credential. Build your own practice material at 200 words per minute, write it into your CAT software with realtime translation on, and grade the raw output without cleanup. For a broader preparation framework, the CRR study guide and CRR training overview go further.
A CRR-Specific Preparation Sequence
Because the credential has a single domain, scheduling is less about topic rotation and more about sequencing the three layers of the skill: dictionary quality, speed and accuracy under test conditions, and procedural fluency. Here is one way to order them across the weeks before a testing window.
Dictionary Audit
- Review prior untranslates and conflicts, and fix entries at the source
- Add vocabulary typical of testimony and Q&A material
- Run short realtime takes below 200 wpm to confirm clean translation
Speed and Sustained Accuracy
- Build to full five-minute takes at 200 wpm, unedited
- Score each take against the 40-error allowance and log error types
- Practice recovery so a single mistake never becomes a 10-word drop
Procedure and Simulation
- Use the exam preview to confirm equipment, wired headphones and Guardian setup
- Rehearse the timers: three minutes for notes, five for the unedited transcript, two for the final paste
- Run full simulations under quiet, camera-ready conditions
This is the only generic scheduling advice you need: put dictionary repair first because it lowers your error floor, then build endurance, then rehearse procedure. If you want a quick-reference version of the key facts, the CRR cheat sheet condenses them onto one page, and our main practice test site offers additional exam-readiness resources.
Eligibility, Fees and Windows
A passing score is useless if you cannot get a seat. Registration is member-only, and eligibility requires current NCRA membership and a current Registered Professional Reporter (RPR) or higher certification. The full prerequisite picture is in the CRR requirements guide.
| Option | Fee | What You Get |
|---|---|---|
| Single attempt | $225 | One attempt at the dictation |
| Dual-test bundle | $338 | Two consecutive attempts in one session |
The dual option is not a longer single test or a written knowledge exam; it gives you two back-to-back attempts in one session. Given the historical pass rates, some candidates find the second attempt valuable, though it is worth deciding in advance whether you will be mentally able to reset between takes. A full cost analysis appears in the CRR certification cost breakdown.
Scheduling follows published alternating registration and testing windows rather than a monthly offering. The published pairing lists registration October 1-20, 2026 with testing November 1-20, 2026, and registration closes at midnight Eastern. Do not assume the test is available every month. The CRR exam dates guide explains how to plan around the windows.
After You Pass: Maintaining the Credential
Earning the CRR is not the end of the obligation. Maintenance requires continued NCRA membership and the underlying RPR or higher credential. The RPR and CRR share one three-year cycle requiring 3.0 total credits, at least 2.0 of which must be continuing education units (CEUs), with up to 1.0 allowed as Professional Development Credits (PDCs). The requirements are not cumulative, so holding both RPR and CRR does not mean 6.0 credits. The detailed rules clarify the credential page's shorthand of 3.0 CEUs.
If you are weighing whether the effort pays off, the ROI analysis, salary guide and CRR jobs overview discuss the career side of the decision.
Frequently Asked Questions
The standard is 96% accuracy on a single five-minute testimony/Q&A dictation at 200 words per minute, transcribed in realtime without editing. The currently linked grading guide permits 40 errors.
No. The CRR has no multiple-choice component and no question count. The realtime dictation is the sole performance component, so preparation should center on timed dictation and unedited output.
Two drops of 10 or more consecutive words, or an uploaded file without printable data. Either condition fails the attempt regardless of the rest of your accuracy.
No. The transcript must be unedited realtime output. The post-dictation timers (three minutes for notes, five for the unedited transcript, two for the final paste) are for submission only, not for editing or additional transcription.
NCRA statistics show 9% across 345 tests in 2023, 8% across 281 in 2024 and 11% across 304 in 2025. These reflect completed tests submitted for scoring, not first-time-candidate rates, and they do not predict any individual's result.