For a trainee revising the GPhC Common Registration Assessment (CRA) who wants volume across both parts, PreRegExamPrep is a reasonable primary bank: it advertises broad coverage of calculations, single-best-answer and extended-matching items with mock exams. Its principal limitation is the one every large bank shares — worked hard and carelessly, it teaches you to recognise its own questions rather than the underlying principles. This workflow is built to stop exactly that: protect unseen items, review misses by error type, and measure transfer on questions you have never seen.
This is a narrow child article that links up to the GPhC decision-tree hub and the iatroX comparison hub; it does not reprint "best bank" rankings or generic price tables. It answers one question: how do you get the volume benefit of PreRegExamPrep without the memorisation trap?
What PreRegExamPrep offers for the GPhC CRA right now
Current-state snapshot. Figures are vendor-reported and last checked on 20 July 2026; confirm anything you are about to pay for on the product page, because contents and prices change.
| Item | What we found (vendor-reported, 20 July 2026) |
|---|---|
| Product type | GPhC/pre-reg exam practice bank with mock exams |
| Headline question count | "2,000+ practice questions" across calculations, SBA and EMQ |
| Components covered | Part 1 (40 calculations) and Part 2 (90 SBA + 30 EMQ) explicitly mapped |
| Mock exams | Yes — "test yourself under exam conditions" |
| Framework alignment | Content mapped to the CRA framework; references Board of Assessors' published feedback (June 2025) |
| AI / adaptive engine | None advertised |
| Guarantee | 7-day money-back guarantee stated |
| Pass-rate marketing | A "94% pass rate" appears in page titles — treat as an unverifiable marketing claim |
| Price / access period | Not clearly disclosed on the pages checked — verify current price and access |
Two honest points. First, this is a genuine CRA bank covering both parts with mocks, so there is no coverage flag. Second, the "94% pass rate" that appears in the site's marketing should be treated with caution: the Board of Assessors does not endorse any commercial question bank, and the provider itself frames its material as practice scaffolding, so no bank can attribute a pass rate to itself in any verifiable way. Use the bank for its volume and mocks, and discount the headline claim.
The exam you are actually preparing for
The GPhC CRA is one day, two parts, both passed in the same sitting (maximum three attempts; verified on pharmacyregulation.org, 20 July 2026):
- Part 1 — Calculations: 40 numerical free-entry questions in 120 minutes.
- Part 2 — Multiple choice: 120 questions in 150 minutes — 90 five-option SBAs and 15 EMQ sets of two (30 EMQ items, eight shared options per set).
The GPhC framework and official sample questions are the standard; PreRegExamPrep is third-party scaffolding aligned to it, not a preview of the paper. Anchor medicines facts to the SmPC/eMC, professional and legal standards to MEP and GPhC standards, and management to NICE and CKS.
Build a blueprint inventory and protect an unseen pool
Before your first block, split the bank in two. Reserve a protected pool of questions — plus the official sample material — that you never study on, keeping them for timed mixed assessment. Study the rest. Then build a blueprint inventory: a row per framework domain and per calculation type, so you can see coverage rather than trust a completion bar. The protected pool is what keeps your later scores honest; without it, every question eventually becomes recognition.
First pass: mixed by default, topic-filtered only where weak
Resist the instinct to grind topic by topic. Topic-filtered blocks cue you — the topic label narrows the answer — so start mixed by default. Use topic filters only where your baseline shows genuinely weak foundations that need concentrated repair; once that domain is above its floor, return it to the mixed pool. Starting mixed trains the cross-domain retrieval the exam actually demands.
Review each miss by error code, not by transcription
The single biggest waste in Q-bank revision is transcribing whole explanations. Instead, tag each miss with a short error code and one corrective action:
- K — knowledge gap (I did not know the fact) → one line to learn, checked against the SmPC/eMC or NICE/CKS.
- A — application error (knew the fact, misapplied it) → one rule for when it applies.
- C — calculation error (method or arithmetic) → redo the calculation cold, note the step that failed.
- P — process error (misread, mis-timed, changed a right answer) → one habit fix.
- D — discrimination error (fell for the distractor) → one "why not the alternative?" note.
A one-line coded action beats a copied paragraph, because it forces you to name what actually went wrong and gives you something specific to retest.
Use transfer practice, not repetition
When you get an item wrong, do not simply re-do that item until you remember it — that trains recognition. Instead, answer a new question that tests the same principle before you ever revisit the original. If you cannot find one in the bank, that is exactly where an unseen bank such as iatroX earns its place: it lets you re-test the principle on an item you have never seen, which is the only honest way to know the learning transferred.
Switch to mixed timed blocks when floors are met
Move to mixed, timed blocks as soon as your domain floors are met — even if your first pass through the bank is incomplete. Completing 100% of a bank is not the goal; getting every domain above its floor and holding accuracy under time is. An incomplete bank with strong, even coverage beats a fully completed one with three neglected domains.
Exit criteria
Stop doing new questions when the data — not the completion percentage — says so: adequate coverage across the framework, stable performance on unseen timed items, correct pacing, retention on retest, and calibration against the official sample material. Any of those still failing is a reason to continue; all of them met is permission to stop and rest before the exam.
A worked seven-day plan (PreRegExamPrep for one job, iatroX for measurement)
Give PreRegExamPrep one job — high-volume drilling and mocks across both parts — and use iatroX for unseen transfer measurement. No proprietary-algorithm claims are involved: you select the domains and run timed, unseen, mixed blocks yourself. This example suits a pre-registration trainee around six to eight weeks out.
| Day | PreRegExamPrep (one job: volume + mocks) | iatroX (unseen measurement) |
|---|---|---|
| Mon | 30-item mixed block; code every miss; one corrective action each | — |
| Tue | Redo yesterday's coded misses via new items testing the same principle | 20-item unseen mixed Part 2 block, timed; record first-pass accuracy |
| Wed | 20 calculations under time; code C-errors; redo cold | — |
| Thu | 30-item mixed block; focus review on A- and D-errors | 20-item unseen mixed block; note the two weakest domains |
| Fri | Topic-filtered repair on the two weak domains only | Retest those two domains, 15 unseen items each |
| Sat | Full-length timed mock (Part 2), strict clock | — |
| Sun | Review the mock by error code; update the blueprint inventory | 30-item unseen mixed block as the week's transfer score |
The iatroX side answers the only question that predicts a pass: on fresh, timed, mixed items, is your accuracy rising? Your PreRegExamPrep percentage is not your exam score — see "Your Q-Bank Percentage Is Not Your Exam Score" — and keep the two banks non-overlapping using the two-Q-bank rule.
Decision checklist: continue, supplement, switch or stop
- Continue if your first-pass accuracy on PreRegExamPrep blocks is rising and its explanations still teach you something.
- Supplement (add unseen measurement) once you have worked most of the bank, your unseen accuracy lags your in-bank percentage by more than about ten points, or a domain remains untested.
- Switch only if sampling shows content out of step with current guidance or the Board of Assessors' published feedback, not because a competitor looks newer.
- Stop buying more input when your limiting factor is pacing or calculation speed — that time belongs in timed mocks and calculations drills.
Base every branch on a measured gap, not novelty or sunk cost.
Frequently asked questions
Is PreRegExamPrep enough for GPhC Common Registration Assessment on its own? For a candidate with sound foundations who works it as a bank rather than a script, it can serve as a primary resource, because it covers both parts and offers mocks. Its weakness is inherent to any single bank: once you have seen its items your rising percentage measures recognition, not understanding, and its marketing pass-rate claim is not something you should treat as evidence. Reserve a protected pool of unseen questions — official sample material plus a measurement bank — so that some of your practice always tests transfer, and it can carry most of the load.
Which GPhC Common Registration Assessment component does PreRegExamPrep not reproduce well? No screen bank fully reproduces the pressure of the free-entry calculations paper, where you type answers under time with no options to recognise, so verify the depth and realism of its calculation practice rather than assuming volume equals readiness. More broadly, a bank reproduces the format of Part 2 well but cannot, by itself, guarantee the standard; only the official sample material does that, which is why calibration against GPhC material remains essential.
How many PreRegExamPrep questions should I complete per day for GPhC Common Registration Assessment? A sustainable target on placement is 30–40 Part 2 items plus a short calculations set daily, always coding and reviewing misses the same day; on study leave, 60–80 with the same discipline. The number matters far less than the review — a coded, retested miss is worth more than ten questions ticked through. Prioritise wrong and low-confidence items over easy repeats, and never let the completion bar set your pace.
When should I stop using PreRegExamPrep and move to mixed mocks? You should be doing mixed blocks from early on, but shift decisively towards full-length timed mocks in the final two to three weeks, once your domain floors are met and blocks stop revealing new gaps. The signal is diminishing returns on fresh items: when new questions surface few genuine errors, your constraint is pacing, stamina and calculation speed, which only mixed timed mocks and drills train.
How should I combine PreRegExamPrep with iatroX without duplicating practice? Give each a distinct role and never test the same item twice. Use PreRegExamPrep for high-volume drilling and mocks, and iatroX for unseen, timed, mixed blocks that measure whether the learning transfers — and specifically for re-testing a missed principle on an item you have never seen. Keep one coded miss log, and treat the iatroX unseen score, not either bank's internal percentage, as your readiness signal. Once an item is recognised, it stops measuring anything.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. Vendor figures (question counts, mock features, guarantees, prices and the marketing pass-rate claim) are vendor-reported as of that date and change without notice — verify them on the PreRegExamPrep product page, and note that the Board of Assessors does not endorse any commercial question bank. Disclosure: iatroX operates a competing UK question bank; its role here is confined to the unseen-MCQ measurement and knowledge layer that PreRegExamPrep's bank does not claim to be. Corrections are welcome via the feedback route on iatrox.com.
References: GPhC CRA structure — pharmacyregulation.org; product details — preregexamprep.com; iatroX GPhC bank, the two-Q-bank rule and "Your Q-Bank Percentage Is Not Your Exam Score".