The CVR Exam Explained: What to Expect and How to Prepare
The CVR Exam Explained: What to Expect and How to Prepare
Fifteen minutes of audio. Three hours and forty-five minutes of transcription.
That ratio is the most useful thing to know about the Certified Verbatim Reporter exam before you register for it. The dictation is short. The transcript production is long. Most candidates who fail did not mishear the words. They ran out of clock, or they lost the segment on commas.
The CVR is the credential granted by the National Verbatim Reporters Association, and it is the certification most voice writers aim for first. The rules below come from NVRA's official CVR, CVR-S, CM and CM-S Certification Rules, Policies, and Procedures manual, revised February 22, 2026.
The Exam Has Two Independent Halves
You must pass both, and they are scored separately.
The Written Knowledge Test is 100 multiple choice questions with a two-hour limit. NVRA weights it as 45% verbatim record, 45% transcript production, and 10% professional responsibilities. You need a minimum score of 70. Answers go on a bubble sheet, so bring a No. 2 pencil.
The Skills Test is three five-minute dictation segments. Each one must be transcribed at 95% accuracy or better. There is no partial credit and no averaging across legs. A leg passes or it does not.
The Three Skills Legs
| Leg | Speed | Length | Approx. words | Errors allowed |
|---|---|---|---|---|
| Literary | 180 WPM | 5 minutes | ~900 | 45 |
| Jury charge | 200 WPM | 5 minutes | ~1,000 | 50 |
| Two-voice Q&A | 225 WPM | 5 minutes | ~1,125 | 56 |
The error allowances are the 95% threshold expressed as raw counts, which is how graders actually work. On the literary leg you get 45 errors. Not 45 misheard words. Forty-five errors of any kind, including a missing comma between a city and a state.
That distinction sinks a lot of otherwise fast candidates. More on that below.
What Happens on Exam Day
All three audio segments run first, back to back. No bathroom breaks between the door closing and the end of the audio portion. Voice writers also face an audibility check: monitors walk the room while you dictate, and if one can hear you from two feet away you get a warning tap on the arm. A tap on the shoulder means you are disqualified for that segment. A second one ends your whole exam.
Then transcription begins. You get 75 minutes per segment, plus ten minutes after each to save your file to a flash drive and hand it to the registrar. You may transcribe the three segments in any order, but only one per timed session, and you cannot start the next one early.
Spell-check, dictionaries, and reference materials are allowed during transcription. Spelling, hyphenation, and compound words are settled by Merriam-Webster's Collegiate Dictionary. Punctuation is settled by The Gregg Reference Manual. You may bring exactly one 3x5 note card.
Formatting is strict and mechanical: one-inch margins on all four sides, Courier New or Times New Roman at 12 point, double spaced, not in all caps. A single-spaced transcript does not get graded at all. Neither does an all-caps one. Your name must not appear anywhere on your test materials, because your six-digit test ID is the only identifier, and a stray header with your name on it is an automatic disqualification.
How Grading Actually Works
NVRA publishes a numbered list of 17 error types that graders apply to your transcript. They cover the obvious categories, wrong words, added words, dropped words, misspellings, wrong names, transposed words, and the ones candidates underestimate: omitted or misplaced Q and A designations, punctuation errors, missing spaces, compound words, capitalization, contractions, verb tense, singular versus plural, acronyms, dates, and numbers.
Punctuation alone is a deep well. An omitted comma in a date costs one error. Missing both commas in that date costs two. City and state works the same way. A semicolon used where the dictation called for a comma is an error. A colon omitted before a list is an error.
NVRA also publishes 14 things that are explicitly not errors, which is worth reading closely because it saves you from over-correcting. The serial comma before a conjunction is optional. Fractions can be words or numerals. Paragraphing is not required. Proper names only have to sound similar, carry the same syllable count, and stay consistent throughout.
One more mechanic that tells you how the exam thinks: the Test Grading Committee stops grading a segment once it finds 15 errors past the maximum allowed. They are not producing a diagnostic report. They are producing a pass or a fail.
What It Costs
The skills test is $185. The Written Knowledge Test is $160 in person, or $200 online. The optional Test Readiness Session is $125. Late registration adds $50. You must be an NVRA member in good standing and hold a high school diploma or equivalent.
Scores behave in your favor if you keep your membership current. A passed skills segment never expires as long as your NVRA membership does not lapse, so you can pass literary in one sitting and Q&A in the next. A passing Written Knowledge Test score is good for three years.
How to Prepare
Speed is the part everyone trains. Accuracy at speed is the part that decides the outcome.
The practical approach is to stop practicing against a vague sense of "that felt pretty good" and start counting errors the way graders count them. Transcribe a five-minute passage, compare it against the reference text word by word, and sort every discrepancy into the error type that caused it. Do that ten times and a pattern shows up. For most people it is not vocabulary. It is commas in dates and city-state pairs, Q and A designations dropped during fast crosstalk, or verb tense drift when the speaker is talking faster than you are thinking.
Then fix that one category before you push the speed up. Practicing at 225 WPM while you are still bleeding four errors a minute on punctuation just gets you fluent at failing.
This is the whole reason VerbatimPrep exists. It scores your practice dictation against the same 17 NVRA error types graders use, tracks your accuracy by error type over time, and runs a speed ladder that opens the next WPM tier only after you hold the accuracy threshold at the current one. Most practice tools on the market were designed for stenographers, not voice writers.
What To Do Next
Download the current CVR Rules, Policies, and Procedures manual directly from nvra.org and read pages 15 through 19, the grading guidelines. Print them. That is your rubric, and almost nobody preparing for this exam has actually read it.
Then start scoring your practice the same way. Start free at verbatimprep.com and get an error-type breakdown on your first session instead of a gut feeling.
Ready to start? Practice voice writing free with real courtroom audio. VerbatimPrep scores your accuracy in real time and coaches you through every mistake. No credit card. No equipment to buy.
Sources: NVRA CVR, CVR-S, CM and CM-S Certification Rules, Policies, and Procedures (rev. February 22, 2026), nvra.org. Confirm current fees, speeds, and grading standards with NVRA before registering.
← All posts