1.3 Scoring, Pass Rate & Test-Day Logistics
Key Takeaways
- CGC scores are reported on a 200–800 scaled score metric; the passing standard is 600.
- Only 170 of 200 items are scored; 30 pretest items do not count but are indistinguishable during the exam.
- ABGC-published first-time pass rates swing widely by administration — 86.06% in August 2025 but 53.62% in February 2025.
- Testing is through Prometric—U.S./Canada test centers or remote proctoring—within ABGC’s published windows.
- Plan pacing for 200 items in 4 hours (~1.2 minutes/item); no calculators are permitted, and scratch paper is issued one two-sheet booklet at a time.
Why scoring and logistics deserve their own study block
Candidates lose points two ways: missing content, and mismanaging the testing session. The CGC uses a scaled score with a fixed passing standard, embeds unscored pretest items, and is delivered exclusively through Prometric under ABGC window rules. Understanding scoring reduces anxiety about “raw percent correct,” and understanding logistics prevents preventable voids (wrong ID, unstable remote setup, arriving for the wrong window).
Scaled scoring and the passing standard
| Scoring element | Official detail |
|---|---|
| Scale | 200–800 scaled score |
| Passing score | 600 |
| Item pool | 200 MCQs; 170 scored + 30 pretest |
| Decision | Pass/fail based on scaled score vs 600, not a simple public “percent correct” cut |
What “scaled” means for your prep
Different forms of the exam can vary slightly in difficulty. ABGC uses psychometric scaling so that a 600 on one form represents the same performance standard as a 600 on another. Practical implications:
- Do not reverse-engineer a casual “I need 70% of 200” rule from the passing score alone—the reported number is a scaled metric anchored at 600.
- Pretest items are used to evaluate future questions; they do not contribute to your scaled score, but you will not know which 30 they are.
- Mark-and-review strategies should be based on clinical reasoning confidence, not on hunting for “experimental” items.
Mental model for the cut score
Treat 600 as the non-negotiable target. Practice exams should be hard enough that you are comfortable with application/analysis stems under time pressure. A practice percent that feels “comfortable” on easy recall decks can still translate to a fail if your weak domains are Domains 3 and 5 under timed conditions.
Pass rates: what public ABGC data suggests
ABGC publishes pass rates by exam form and cohort, and the numbers are far more volatile than any single headline figure suggests. These are ABGC's own published results:
| Administration | First-time | Repeat | Total (all testers) |
|---|---|---|---|
| August 2021 | 79.9% (n=344) | 57.1% (n=28) | 78.3% (n=373) |
| February 2022 | 75.9% (n=116) | 70.9% (n=86) | 73.8% (n=202) |
| August 2022 | 84.0% (n=374) | 68.8% (n=48) | 82.2% (n=422) |
| February 2023 | 73.5% (n=98) | 71.4% (n=70) | 72.6% (n=168) |
| August 2023 | 69.6% (n=402) | 56.6% (n=90) | 67.2% (n=492) |
| March 2024 | 69.9% (n=143) | 57.6% (n=118) | 64.3% (n=261) |
| August 2024 | 79.6% (n=418) | 57.5% (n=73) | 76.3% (n=491) |
| February 2025 | 53.6% (n=138) | 40.8% (n=98) | 48.3% (n=236) |
| August 2025 | 86.1% (n=495) | 63.5% (n=115) | 81.8% (n=610) |
Three things this table actually tells you
- Do not quote a single "~80% pass rate." Published first-time results run from 53.6% (February 2025) to 86.1% (August 2025). A number you saw on a forum is probably the total rate for one window, not the first-time rate — 81.8% for August 2025 is the all-testers figure, while first-time takers passed at 86.1%.
- August and February are different animals. August cohorts are large (400–600 testers) and dominated by new graduates fresh out of an ACGC program; February cohorts are small (100–250) and carry a much higher proportion of repeat and off-cycle testers. February totals have run below August totals in every year shown. ABGC states both forms are built to the same specifications and equated to the same standard, so the gap reflects cohort composition, not an easier August exam.
- Repeat testers pass at markedly lower rates — 40.8% to 71.4% across these windows, and below 65% in six of nine. Retaking without changing your study method is a losing strategy; a failed attempt returns a diagnostic score report with High/Medium/Low performance bands by subdomain, and that report is the thing to build the next plan around.
| Interpretation | Study implication |
|---|---|
| First-time results span 53.6%–86.1% | Cohort context matters; do not treat any one figure as your personal odds |
| February windows historically underperform August | If you test in February, assume less peer momentum and build a stricter schedule |
| Repeat rates cluster in the 40s–60s | Plan a genuinely different approach after a fail; budget the $500 repeat fee and the emotional load |
| Rates are published per window, not per candidate | Compare yourself to timed, mixed-domain practice performance instead |
The February 2025 result — 53.6% first-time, 48.3% overall — drew enough attention from the genetic counseling community that the ABGC Board issued a public update on exam outcomes in December 2025 and announced a refreshed practice exam for June 2026. Treat that window as evidence that a weak administration is survivable but real: prepare for the harder scenario, not the friendliest published number.
Test delivery: Prometric center or remote
ABGC delivers the CGC exam through Prometric:
- In-person at Prometric test centers in the United States and Canada
- Remote proctoring (online), subject to Prometric/ABGC environment and equipment rules for that administration
Choosing a modality
| Modality | Advantages | Watch-outs |
|---|---|---|
| Test center | Controlled environment, fewer home-tech variables | Travel time, seat scarcity in some cities, center rules on personal items |
| Remote proctoring | No commute; test at home/office if approved | Strict room scan, webcam/mic requirements, internet stability, interruption policies |
If you choose remote, complete a system check days before the appointment. A failed webcam, shared housing noise, or dual-monitor setup that violates rules can cost you the seat.
Windows and scheduling logistics
You sit during an ABGC-authorized window after your application is approved. Planning anchors for this guide:
| Window | Dates |
|---|---|
| 2026 summer window | August 1–31, 2026 |
| 2027 winter window | February 1–28, 2027 |
After authorization, schedule promptly. Popular center times and remote slots fill; waiting until the last week of the window increases risk.
Pacing and in-exam strategy (4 hours / 200 items)
| Metric | Value |
|---|---|
| Total time | 4 hours (240 minutes) |
| Items | 200 |
| Average | ~1.2 minutes per item |
Practical pacing approach
- First pass: Answer every item you can decide in under ~90 seconds; flag calculation-heavy Bayesian or multi-step counseling ethics items.
- Second pass: Return to flagged stems with remaining time.
- Never leave blanks if the interface allows review—there is no penalty reduction for guessing beyond opportunity cost of time.
- For long clinical stems, read the lead-in question first, then scan the vignette for the data that question actually needs (inheritance, gestational age, variant classification, counseling goal, legal constraint).
Content-linked timing traps
- Domain 2 risk math can consume disproportionate time—and you are doing it by hand, since no calculator is allowed. Set a personal cap (e.g., two minutes) before flagging.
- Domain 4 counseling stems often have two “nice” answers; spend seconds on which response best fits the model/ethics, not on rewriting the vignette in your head.
- Domain 5 legal items are often shorter stems—bank time there for harder clinical items.
Test-day requirements (high-yield)
Exact Prometric rules are confirmed in your appointment confirmation and the ABGC Candidate Guide (October 2025 edition for current policy language). Core expectations candidates must plan for:
| Rule | ABGC's published policy |
|---|---|
| Identification | One valid, government-issued photo ID with signature, in Latin characters, matching your registration. Testing outside your country of citizenship requires a passport |
| Arrival | 30 minutes before the appointment (test center) or 30 minutes before launch (remote). Arriving late means you are not allowed to test and forfeit the fee |
| Calculators | No calculators are permitted — none, in either modality |
| Scratch paper | Pencils are provided; you receive one booklet (two sheets) at a time and may exchange it as often as needed, but all scratch paper must be returned or no score report is issued |
| Personal items | Everything goes in a locker; at a test center you will be asked to empty your pockets. A phone or alarm heard during the exam means dismissal and a forfeited administration |
| Breaks | You may break whenever you wish, but no additional time is given for it |
| Guessing | There is no penalty for guessing — answer every item before ending the exam |
| Remote setup | Clear surface holding only the computer, a chair, and the computer itself; install ProProctor and run a System Check before exam day |
The no-calculator rule deserves its own line in your study plan. Domain 2A puts Bayesian tables, carrier-frequency multiplication, and Hardy-Weinberg arithmetic on a paper-and-pencil footing. Practise those calculations by hand, in fractions, on two sheets of scratch paper — not with a phone calculator beside your practice bank.
Retakes, reschedules, and refunds
| Policy | Rule |
|---|---|
| Number of attempts | ABGC does not limit the number of attempts, but allows only one attempt per testing window |
| Reschedule / cancel | Free if changed at least 30 days before the appointment; a fee applies within 29 days; cannot be rescheduled or cancelled within 5 days |
| Rollover to another window | $75 rollover fee when you meet ABGC's criteria for the change |
| Refund | Requested by email to ABGC after cancelling the appointment; approved refunds are paid minus a $125 processing fee; no refund for a no-show |
| Failure to schedule / report | Forfeits the attempt and the fee; you must re-meet all eligibility requirements and pay again for the next cycle |
| Score turnaround | Scores are normally reported immediately at the end of the session; during the Job Analysis period (every 5 years — next practice analysis in 2027) one window will not offer immediate scoring |
Day-of failure modes (avoidable)
- Name mismatch between ABGC registration and ID
- Attempting remote testing from a café or shared dorm room
- Assuming break policies match another board exam you took (follow CGC/Prometric rules for this appointment)
- Spending 10 minutes fighting with a flagged calculator-style risk item while unanswered easy Domain 1 items remain
Connecting logistics back to the blueprint
You have 170 scored items distributed across five domains (34 / 32 / 35 / 30 / 39). On test day, that distribution is invisible item-by-item—your job is consistent accuracy under the 600 scaled-score standard. Use practice blocks that mix domains and cognitive levels so the Prometric session feels familiar.
Official references for scoring and administration
- ABGC Certification Process, Exam and Fees
- ABGC Candidate Guide (October 2025) — administration policy, restrictions, and pass/fail methodology
- ABGC published pass rates by exam form and cohort — the only source that separates first-time from repeat rates
- ABGC Program Directors Toolkit — cycle dates, fees, and score-report samples
- Prometric ABGC exam information pages for scheduling and system checks
What scaled score is required to pass the CGC exam?
ABGC reports an August 2025 total pass rate of 81.80% and a first-time pass rate of 86.06%. A candidate cites "the August 2025 pass rate was about 82% for first-time takers." What is the error?
Who delivers the CGC examination for ABGC?
A candidate plans to work Bayesian carrier-risk tables on the CGC exam using a handheld calculator and a stack of scratch paper kept at the desk. Which statement reflects ABGC and Prometric policy?
Which statement about CGC scoring and pretest items is correct?