There is no single best Prescribing Safety Assessment (PSA) resource, and any article that names one is selling something. The right resource depends on your profile — first attempt or retake, how many weeks you have, your budget, and whether your weakness is knowledge, construction, pacing or calibration. This decision tree routes you to the smallest stack that fixes your actual gap, using the official material as the fixed calibration anchor and a bank for the volume it cannot provide. It is the exam-level hub for this decision; platform-specific audits sit beneath it.
Start by segmenting yourself
The branch you follow depends on which of these you are. Most candidates are a blend; pick the dominant one.
- First attempt, average preparation: you need coverage and calibration, not exotic tools.
- Retake: something specific failed last time — diagnose it before buying anything new.
- Final-year student or F1 with a broad base: knowledge is largely there; the risk is construction and pacing.
- Weak foundations: you need teaching before testing — questions alone will demoralise, not build.
- Strong knowledge, poor pacing: you know the medicine but run out of time — you need timed papers.
- Strong on MCQs, weak on practical prescribing: you pick the right answer but cannot construct a prescription from a blank field — the most common hidden failure mode.
Define the minimum stack
Almost every candidate needs the same four-slot stack, and no more. Adding resources past this dilutes attention and duplicates content.
- One primary question bank for volume and recognition breadth.
- The official PSA practice papers as the non-negotiable calibration standard — free, and the closest thing to the real exam.
- One teaching or reference source, only where foundations are weak — used to build a topic, then set aside.
- One modality tool where relevant — blank-field construction practice, or timed-mock software — chosen to fix your specific gap.
If you already have a bank and the official papers and you are practising construction, you likely do not need to buy anything else. The content-gap analysis of what MCQ practice cannot train tells you which modality tool, if any, your profile actually needs.
Budget bands
Verify every price on the day you buy; the figures below are indicative and vendor-reported where named.
- Free / low-cost: the official PSA practice papers (free) plus the free iatroX PSA bank cover calibration and volume at no cost — a legitimate complete-enough stack for a well-prepared candidate. A low-cost recognition bank such as Geeky Medics (vendor-reported £19.99/year, 20 July 2026) adds mocks and volume cheaply.
- One premium resource: if you buy one paid tool, buy the one that fixes your gap — a bank with strong mocks for volume and pacing, or a construction-focused resource if that is your weakness. Do not buy two banks.
- Comprehensive stack: a primary bank, the official papers, a teaching source and a construction tool. Justify each slot against a named gap; if you cannot, drop it.
Time bands — and what to omit
The honest planning skill is subtraction. State what you will not do.
- Under 4 weeks: official practice papers first for calibration, then targeted volume on your weakest high-mark areas (Prescribing above all). Omit comprehensive topic-by-topic revision and any second bank; there is no time.
- 4–12 weeks: a full first pass on one bank, weekly timed mocks, and dedicated construction practice. Omit a third resource and any tool that does not map to a logged gap.
- More than 12 weeks: build foundations early with a teaching source, then transition to volume and timed mocks, protecting an unseen pool for the final fortnight. Omit the temptation to hoard resources; more time is not a reason to buy more tools.
Decision matrix: platforms mapped to their best job
Match the tool to the job, not to a ranking. Verify current counts and prices on each product page; qualitative "best job" mappings below avoid quoting figures we have not confirmed.
| Resource | Best job | Weakest for |
|---|---|---|
| Official PSA practice papers | Calibration — the gold standard for format and standard | Volume (finite set) |
| Geeky Medics PSA bank + mocks | Low-cost recognition volume and timed mocks | Blank-field construction |
| A large premium PSA bank (e.g. BMJ OnExamination, Pastest — verify) | Depth, explanations and analytics | Real construction under exam tolerance |
| A dedicated calculation/QUPI-style tool (verify) | Calculation drilling | Whole-exam realism |
| iatroX PSA bank | Free unseen, timed measurement and knowledge grounding | It is not a construction simulator |
| A prescriber / clinical pharmacist | Marking constructed prescriptions; high-risk judgement | Volume and availability |
The pattern to notice: no single product does every job. The official papers calibrate but cannot supply volume; banks supply volume but under-train construction; only a human reliably marks a constructed prescription against the exam's tolerance. Your stack should cover jobs, not accumulate brands.
Read your baseline, then follow one branch
The decision tree only works if you feed it real data about yourself rather than a hunch. Sit one fresh, timed baseline before you choose a branch, and read it along four axes rather than as a single percentage. First, your domain profile: which of the seven clinical domains fell below the rest, and is any high-mark Prescribing item among them. Second, your question-type profile: is the weakness concentrated in Prescribing and Calculation Skills — the construction-heavy types — or spread evenly across the recognition types. Third, your pacing: did you finish all 60 items, and did the Prescribing section get the time its marks deserve. Fourth, your error character: are your mistakes knowledge gaps, misreads, or safety-critical omissions, because those three route to different tools.
A candidate whose baseline shows even domain coverage, high recognition accuracy and a rushed Prescribing section is a pacing problem — Profile B — and should buy timed papers, not a new bank. A candidate with strong percentages but omissions on constructed prescriptions is Profile C's construction problem and needs a prescriber's marking, not more volume. The baseline is what tells you which branch you are on; without it, you are guessing, and guessing usually defaults to buying more questions you do not need.
For a retake, the baseline is diagnostic gold: compare it against your recollection of what failed last time. If the same domain or the same question type fails again, the resource you used did not fix it, and repeating it is sunk-cost reasoning — change the modality, not just the brand. If a new area has slipped while an old one recovered, your coverage has drifted and you need breadth, not more depth in what you already know. Either way, let this year's baseline, not last year's plan, choose the branch.
Cannibalisation guardrail
This hub deliberately keeps platform descriptions short. For the detailed evidence on any one product — its live counts, prices, mock structure and limitations — follow the narrow child audits (for example, the Geeky Medics PSA workflow) rather than expecting the full case here. Summarise once, link down for depth, and avoid repeating long platform write-ups across articles; that keeps each product's detail in one authoritative place and this article focused on the decision.
Worked examples: three profiles
Profile A — "Priya", final-year student, 6 weeks, strong knowledge, never constructs prescriptions. Stack: official papers + free iatroX bank + a construction routine. Weekly allocation: two official/mixed timed papers, ten blank-field prescriptions a day marked against the official criteria (a peer prescriber checks two), and a short unseen transfer block. Exit criteria: 9/10 prescriptions safe as written, no wrong-route errors on high-risk drugs, both papers finished with review time.
Profile B — "Sam", F1 retake, 4 weeks, failed last time on pacing. Diagnose first: the gap is time, not knowledge. Stack: timed mocks + the official papers, nothing new to learn. Weekly allocation: three full timed papers under a per-section clock, a timing log, and Prescribing-first sequencing drilled deliberately. Omit new topic revision. Exit criteria: all 60 items attempted, Prescribing section never rushed, stable unseen timed score.
Profile C — "Aisha", first attempt, 12 weeks, weak foundations in therapeutics. Stack: a teaching/reference source early, then one bank, the official papers, and a construction routine. Allocation: weeks 1–4 build topics and calculation method; weeks 5–9 first-pass volume plus weekly mocks; weeks 10–12 unseen protected papers and construction sign-off. Exit criteria: coverage matrix complete, calibration against official papers, no recurrent safety-critical omission.
Evidence hierarchy
When sources conflict, rank them: official PSA material first for format and standard; primary guidance (the SmPC/eMC, NICE and CKS) for clinical content; vendor pages for product facts, always labelled vendor-reported; and independent user testing for experience. Never let a vendor's claim override the official blueprint, and never let a bank's explanation override the SmPC/eMC on a medicines fact. Your percentage on any bank is a coverage signal, not a score — why is explained here.
Frequently asked questions
How do I know whether I have covered the full Prescribing Safety Assessment blueprint? Build a matrix of the eight question types against the seven clinical domains and mark the cells you have practised on unseen items, weighting Prescribing and the recurring high-risk drugs most heavily. You have covered the blueprint when no domain row is empty, no high-mark question type is untouched, and you have confirmed dose accuracy on anticoagulants, insulin, opioids and aminoglycosides — regardless of your overall percentage. A high completion figure with an empty paediatric or O&G row is not coverage.
Can one question bank be enough for Prescribing Safety Assessment? One bank can be your only bank, but it cannot be your only preparation, because the official practice papers are the calibration standard and construction needs blank-field practice a bank does not provide. The minimum is a bank for volume, the official papers for calibration, and construction reps marked to the official criteria. Add a second bank only to measure transfer on unseen items, never to repeat content.
What should I measure instead of my overall Q-bank percentage for Prescribing Safety Assessment? Measure your unseen, timed score; your prescription-construction accuracy scored against the official criteria; your dose, route and frequency error rate on high-risk drugs; and your completion under time. These four predict readiness; the overall percentage, especially on re-seen questions, does not, and can rise while your unseen performance falls.
When should I stop doing new Prescribing Safety Assessment questions? Stop when new questions stop changing your error profile — when a fresh unseen block produces only errors you have already logged. At that point switch your remaining time to timed full papers, construction reps on weak domains, and official-material calibration. More volume past this point inflates the percentage without improving readiness.
Which Prescribing Safety Assessment resource should I use for my weakest component? Route by the deficit: calculation weakness needs focused calculation drilling with every figure checked against the SmPC/eMC; construction weakness needs blank-field practice marked by a prescriber or against the official criteria; formulary-navigation weakness needs timed find-the-fact drills in the on-screen resources; and pacing weakness needs timed full papers, nothing else. Use this decision tree's profiles to pick the single resource that fixes your dominant gap rather than buying breadth you do not need.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. Named product figures — for example the Geeky Medics £19.99 annual price — are vendor-reported and may change; verify every price and count on the product page on the day you buy. Other platforms named (BMJ OnExamination, Pastest, QUPI-style tools) are illustrative of a job, not endorsements, and their current specifics should be confirmed on their own pages. The PSA format and mark weightings are taken from the official PSA blueprint and practice papers. Medicines facts should be checked against the SmPC/eMC, with NICE and CKS for practice guidance.
Disclosure: iatroX operates a PSA question bank and competes with several resources named here. We have kept iatroX's role to the jobs a recognition bank does not claim — free unseen, timed measurement and knowledge grounding — and have named plainly where the official material or a human prescriber is the right tool instead. Corrections are welcome via the feedback route on iatrox.com.
References: PSA blueprint and practice papers (prescribingsafetyassessment.ac.uk and the BPS Assessment learner portal); MSC Assessment PSA guidance; vendor product pages for the platforms named; SmPC/eMC and NICE/CKS for content. Internal: the iatroX PSA bank, the comparison hub, and Your Q-Bank Percentage Is Not Your Exam Score.
Complete a fresh Prescribing Safety Assessment baseline in iatroX →