Trainee pharmacists preparing for the Common Registration Assessment deserve the format in exact numbers, because the structure itself dictates the preparation. Part 1 consists of 40 calculation questions over 120 minutes. Part 2 consists of 120 questions, 90 single best answer and 30 extended matching, over 150 minutes. Two papers, two different examinations in everything but the certificate, and the commonest preparation error is treating them as one subject with two sittings. Format as published by the General Pharmaceutical Council, checked 29 August 2026.
Part 1: what 40 questions in 120 minutes rewards
Three minutes per calculation sounds generous and is the paper's central deception: the budget must absorb reading, extraction, method selection, computation, unit discipline and checking, and a single derailed question quietly taxes five others. The paper rewards a specific trained state: method fluency across the calculation families, doses and concentrations, dilutions, displacement values, infusion rates, unit conversions, molecular calculations, so deep that selection is instant and execution is mechanical; free-entry discipline, because answers are produced, not recognised, which removes the multiple-choice safety net and punishes sloppy arithmetic that option-elimination would have caught; and checking as a built habit, the sanity pass on magnitude and units that catches the transposition before it costs a mark. Training implication: volume practice in genuine free-entry format under time, not multiple-choice calculation questions, which rehearse the wrong cognitive act.
Part 2: the 90/30 split and the 150-minute clock
Seventy-five seconds per question, blended across two formats that think differently. The 90 SBAs are ranking exercises, several defensible options, one best, where the skill is systematic elimination and best-available judgement across clinical pharmacy, therapeutics, law and ethics, and where reading speed under pressure is a trainable variable candidates neglect. The 30 EMQs are matching exercises against option lists, rewarding pattern recognition across themes and punishing candidates who re-read the full option list for every item instead of working from the stem outward. The mixed clock is its own skill: pacing that banks time on confident SBAs to spend on EMQ themes, or the reverse, according to your own profile, which only mixed timed practice reveals. Preparation should therefore rehearse the exact composition, and mock papers reproducing the true 90-plus-30 split at true length are worth more than larger volumes of unstructured questions; when judging any bank's mocks, that fidelity is the first thing to check, along with the counting and pricing disciplines that apply across the category: /blog/what-does-5000-questions-mean-qbank-counting.
The reference-handling skill the format quietly tests
The assessment is built around safe practice with the resources a pharmacist actually uses, and preparation should treat permitted-reference discipline as examinable: knowing the current rules on what is available in each paper, and training the retrieval habit, when to look up, how to navigate fast, when looking up is slower than knowing, under the same clock as everything else. Candidates who practise only closed-book, or only with leisurely reference use, arrive with the right knowledge and the wrong reflexes; the trained state is knowing cold what must be known cold, and retrieving cleanly what may be retrieved. Check the GPhC's current guidance on permitted resources directly in your final month, since the rules are the examining body's to state and yours to rehearse.
Frequently asked questions
Which part fails more candidates?
Sitting-level figures vary with cohorts, and the strategic answer is stable: Part 1's free-entry calculations are the more unforgiving format per question, and candidates who under-train them because the paper is shorter are making the classic error.
How early should timed practice start?
From the beginning for calculations, in small daily sets, because fluency is a volume product; full mixed Part 2 papers belong in the final two months, at true composition and length, at least twice.
Do university-style MCQs prepare you for Part 2?
Partially: content transfers, format does not; the SBA ranking discipline and EMQ list navigation are their own skills, and the assessment's composition is exact enough that practice should mirror it exactly.
How many full mocks should I sit before the assessment?
At least two full Part 2 papers at the true 90-plus-30 composition and 150-minute clock, and weekly timed Part 1 calculation sets from early in preparation; stability across them, not a single flattering score, is the readiness signal, per the calibration logic at /blog/why-qbank-percentages-are-not-comparable.
Is the June or later sitting easier?
Cohort composition explains sitting-to-sitting variation better than paper difficulty, and the regulator's own reporting points the same way; choose your sitting by readiness and logistics, not folklore.
What single weekly habit best serves Part 1?
A fixed timed set of mixed calculation families in free-entry format, marked immediately and error-logged by family, because the log converts scattered mistakes into a target list, and the families you avoid are reliably the ones the paper finds.