How many MRCPCH questions should you do each day?

There is no evidence-based universal daily MRCPCH question target. Choose a volume that leaves enough time to answer without notes, review every important error and uncertain success, record the cause, and retrieve the corrected principle again later. Start from available focused time rather than an impressive question count. Increase volume only while review quality, syllabus coverage, delayed retention and recovery remain stable; reduce it when explanations become superficial or the same errors recur.

"How many questions should I do?" sounds numerical, but the useful answer depends on what one completed question produces. Ten carefully reviewed first-seen questions may expose and repair more learning gaps than fifty familiar questions clicked through rapidly. Conversely, a candidate who never progresses beyond tiny untimed sets may not develop coverage or paper-level endurance.

The goal is neither maximum volume nor perfect analysis of every option. It is the largest sustainable question workload that still completes the learning loop.

This article provides non-clinical MRCPCH study-planning guidance. It does not set an RCPCH requirement, endorse a commercial question bank or predict a pass.

RCPCH does not prescribe a daily question target

As checked on 24 September 2026, RCPCH describes the current theory-examination structure and syllabi but does not publish a required number of practice questions per day.

The live structure page currently states:

Component	Current examination structure	Implication for practice
FOP	100 best-of-five single best answer questions	Full-paper rehearsal eventually needs sustained pacing across 100 items
TAS	100 best-of-five single best answer questions	Volume must not replace science-mechanism review
AKP	120 best-of-five questions across two 60-question papers	Practice should eventually include separate paper-level pacing and the between-paper reset

These are examination structures, not daily revision prescriptions. The official number of questions in a paper does not mean candidates should complete that number every day.

Use the live RCPCH structure and syllabi page for current format claims. Formats can change after this article is updated.

Why raw question counts are a weak learning measure

A question-bank counter measures completed items. It does not reveal whether you:

• answered from memory or looked up the topic;
• recognised a repeated item;
• distinguished the best option from a merely plausible one;
• read and understood feedback;
• corrected the reason for the error;
• retained the correction after a delay; or
• transferred it to a different question.

Two candidates can both record "40 questions" while completing very different learning work.

Question volume is an input. The useful output is a corrected decision that remains retrievable in a new context.

Research in health-professions education supports retrieval practice and distributed practice, while also showing substantial variation in intervention design. Feedback research generally favours structured, usable feedback over no or unstructured feedback. None of this research establishes a universal MRCPCH daily number.

The question-bank study method explains the attempt - explain - review - retrieve loop in detail.

Use the PACE workload framework

PACE is a PaedVault editorial framework, not an RCPCH rule or validated score predictor:

1. P - Purpose of the session
2. A - Available focused time
3. C - Correction cost
4. E - Evidence after a delay

A daily target is defensible only when all four are visible.

P - Define the purpose before the number

Different sessions need different volumes.

Learning sessions

Purpose: expose gaps, build reasoning and use feedback.

Typical features:

• mainly first-seen questions;
• explanation before revealing the answer;
• detailed review of errors and low-confidence successes;
• reference checking where a time-sensitive claim matters; and
• a small number of delayed reattempts scheduled.

The completed count may be lower because each question carries a larger correction cost.

Retrieval-maintenance sessions

Purpose: keep previously studied material accessible.

Typical features:

• short mixed sets;
• limited note use;
• faster review when reasoning is genuinely secure;
• delayed items selected from earlier errors; and
• attention to whether familiarity, rather than retrieval, produced the answer.

Coverage-sampling sessions

Purpose: test whether large syllabus areas remain unexamined.

Typical features:

• broad topic mix;
• first-seen items where possible;
• error causes tagged by syllabus area; and
• follow-up study scheduled outside the question block.

Timed-performance sessions

Purpose: rehearse completion, pacing and recovery under representative conditions.

Typical features:

• an uninterrupted block or full paper;
• no answer review until the timed block ends;
• paper-level checkpoints;
• a separate review session; and
• completion and process data recorded alongside accuracy.

The mock-exam strategy guide explains why a mock is a diagnostic event, not just a large daily target.

A - Start from available focused time

Set the time boundary first. Then subtract the time needed for review.

For example, a candidate with 60 focused minutes might reserve:

• 25 minutes for first-seen questions;
• 25 minutes for reviewing decisions and explanations; and
• 10 minutes for recording corrections and scheduling delayed retrieval.

The resulting question count is whatever fits those standards. It is not decided in advance by copying another candidate.

Use a completion equation

A simple planning equation is:

Daily capacity = focused minutes ÷ average complete-loop minutes per question

"Complete-loop minutes" includes attempting, explaining, reviewing and recording the next action - not merely clicking an option.

Measure the average across three ordinary sessions. Do not estimate it from your fastest block.

Adjust for session mode

Session mode	Expected relative volume	Review depth
New weak topic	Lower	High
Mixed first-seen block	Moderate	High for errors and uncertainty
Delayed retrieval	Moderate	Targeted
Timed paper	High during attempt	Deferred but still required
Familiar repeat set	Potentially high	Interpret cautiously

The table does not prescribe exact numbers. It shows why one fixed daily quota cannot fit every session.

The rota-based revision guide can help match session size to different day types without pretending every day has equal capacity.

C - Measure correction cost

An error is not "reviewed" because the explanation was opened.

Classify the cause:

• missing knowledge;
• failed retrieval;
• stem misreading;
• poor option comparison;
• outdated or conflicting source;
• time pressure;
• overconfident guess; or
• answer changed without new evidence.

Then choose the smallest repair that addresses that cause.

Error cause	Suitable next action
Missing knowledge	Learn the bounded principle from an authoritative source
Failed retrieval	Schedule a delayed cue without reopening the full explanation
Stem misreading	Restate what the question asks before comparing options
Option confusion	Write the discriminating feature between the leading choices
Time pressure	Retest in a timed mixed block after reasoning is corrected
High-confidence error	Prioritise review because the internal model was strongly wrong

If today's target creates more corrections than can be processed, the target is too high for that session mode.

The incorrect-answer review guide provides a cause-based correction workflow.

Review uncertain correct answers

A correct answer can still be a weak learning outcome when it came from guessing, an invalid rule or recognition of a repeated item. Mark confidence before revealing the answer, then review:

• incorrect answers;
• correct low-confidence answers; and
• correct high-confidence answers with flawed reasoning.

Do not spend equal time on every correct answer. Use uncertainty and reasoning quality to triage.

Stop collecting unresolved debt

Unresolved review debt includes:

• explanations saved but not read;
• errors logged without a repair;
• flashcards created but never retrieved;
• references opened without a decision; and
• repeated mistakes left as "careless".

A growing debt is evidence to reduce new-question volume temporarily. Clearing it is not losing momentum; it completes work already started.

E - Demand evidence after a delay

Immediate understanding can feel convincing because the explanation is still visible. Learning becomes more credible when the corrected principle can be retrieved later and applied to a different cue.

Use three evidence levels:

1. Immediate correction: Can I explain why the original decision failed?
2. Delayed retrieval: Can I produce the principle after a gap without reopening the explanation?
3. Transfer: Can I use it in a fresh question with different surface details?

The spaced-repetition guide explains how to schedule delayed retrieval without turning every explanation into a permanent card.

Keep new and repeat questions separate

Label each item as:

• first-seen;
• delayed reattempt;
• recent repeat; or
• recognised/familiar.

Daily totals that mix these categories can mislead. A day of 60 familiar repeats is not equivalent to 60 first-seen questions, and first-seen performance should not be compared directly with a correction set.

Use a weekly evidence review

At the end of the week, ask:

• How many first-seen questions were completed?
• Which syllabus areas remained unsampled?
• What proportion of errors received a defined repair?
• Which corrections survived delayed retrieval?
• Did repeated errors reduce?
• Did completion and review quality remain stable as volume changed?

The goal is to adjust the next week, not to reward the largest total.

How to find your sustainable daily range

Use a two-week calibration instead of choosing one permanent number.

Step 1: measure three normal sessions

For each session, record:

• focused minutes;
• first-seen questions attempted;
• delayed reattempts;
• errors and uncertain correct answers;
• review completed; and
• unresolved review debt.

Calculate the observed complete-loop pace.

Step 2: set a range, not a single quota

Create three targets:

• minimum useful session: small enough for a difficult rota day but still includes review;
• standard session: fits an ordinary focused day; and
• extended session: reserved for planned long blocks or timed rehearsal.

For illustration only:

Day type	Attempt plan	Completion condition
Short workday session	One small mixed set	Important errors reviewed and next retrieval scheduled
Standard study day	Two first-seen blocks	Review quality remains stable across both
Long rehearsal day	Component-sized timed block	Separate review time is protected afterwards

Do not copy an illustrative block size from someone else. Measure your own complete-loop pace.

Step 3: raise volume experimentally

Increase the target modestly for several comparable sessions. Keep topic mix, item novelty and review rules similar.

Retain the increase only if:

• review completion remains high;
• recurring errors do not rise;
• delayed retrieval stays stable;
• coverage broadens;
• sleep and ordinary recovery are not being sacrificed; and
• the next session begins without substantial unresolved debt.

Step 4: reduce volume when quality breaks

Reduce or restructure the target when:

• explanations are skimmed to protect the count;
• guesses are not marked;
• the same cause repeats without repair;
• fresh-question accuracy falls while familiar repeats dominate;
• notes and cards accumulate without retrieval;
• timed completion is practised before reasoning is stable; or
• fatigue makes the final block materially less reviewable.

The revision-fatigue guide helps distinguish a demanding session from a study system that is no longer recovering.

How volume should change across preparation

Early phase: sample broadly

Use moderate first-seen sets across the syllabus. The main goal is to discover coverage and reasoning gaps, so correction time is substantial.

Middle phase: expand mixed retrieval

As foundations become more stable, increase mixed sets and delayed retrieval. Protect time for high-confidence errors and cross-topic discrimination.

Late phase: add paper-level work

Use current official format and timing for representative rehearsals. A large timed block creates a separate review obligation; do not hide that cost by counting only attempt time.

Final days: stop chasing a lifetime total

A last-minute question-count sprint can create unresolved feedback, sleep loss and unnecessary uncertainty. Use bounded sets to maintain process, then follow the practical closure plan in the final 72-hours guide.

Common question-volume mistakes

Copying a social-media number

The number lacks your rota, baseline, question difficulty, novelty, review rules and exam runway.

Counting repeated questions as fresh evidence

Repeats can support delayed retrieval, but familiarity may inflate speed and accuracy. Label them separately.

Reviewing only wrong answers

Low-confidence correct answers and lucky guesses can reveal fragile reasoning.

Turning every explanation into notes

Create a durable item only when it has a future retrieval purpose. Otherwise the note backlog can consume the time that questions were meant to protect.

Increasing volume during a plateau without diagnosis

More of the same process may reproduce the same result. Use the score-plateau guide to identify the limiting mechanism first.

Treating a completed question bank as syllabus mastery

Completion does not establish coverage quality, retention or transfer. Map evidence against the current syllabus using the syllabus revision guide.

A one-page PACE tracker

Field	Daily entry
Session purpose	Learning, maintenance, coverage or timed performance
Focused minutes	Planned and actual
First-seen questions	Count
Delayed reattempts	Count
Important errors/uncertainty	Count and cause
Review completed	Yes, partial or no
New review debt	Specific items
Delayed evidence	What survived and what failed
Coverage gap found	Syllabus area
Next volume decision	Maintain, increase, reduce or change mode

Review the tracker weekly. If recording it takes longer than the learning decision it supports, simplify it.

Final answer

There is no universal number of MRCPCH questions that every candidate should complete each day. Set the session purpose, protect focused time for review, measure the cost of correcting errors and require evidence after a delay. The right range is the largest workload that preserves those conditions across ordinary weeks.

Use PACE: purpose, available time, correction cost and evidence. Increase volume only as an experiment with stable review and retention. When the count begins to replace thinking, feedback or recovery, reduce it - even if the dashboard looks less impressive.

Frequently asked questions

Is 100 MRCPCH questions a day too many?

It may be appropriate for an occasional component-sized timed rehearsal, but it is not an official or universal daily target. The answer depends on question novelty, available time and whether review is completed afterwards. If attempting 100 questions leaves important errors, uncertain successes and feedback unresolved, the learning target was too large for that day.

Is 20 MRCPCH questions a day enough?

It can be enough for a useful session when the questions are first-seen, attempted without notes and followed by high-quality review and delayed retrieval. Across the whole preparation period, also check whether that pace can provide broad syllabus sampling and eventual timed-paper rehearsal before the examination.

Should I review every MRCPCH question explanation?

Review every important error, low-confidence correct answer and answer reached through flawed reasoning. Secure correct answers may need only a brief check. The aim is not identical review time for every item; it is to detect and repair decisions that are wrong, fragile or falsely confident.

Do repeated MRCPCH questions count towards my daily total?

Track them, but label them separately from first-seen questions. Delayed reattempts can test whether a correction remains retrievable, while recent or recognised repeats may inflate speed and accuracy through familiarity. Do not combine the categories when judging coverage or readiness.

When should I increase my daily MRCPCH question target?

Increase it modestly after several comparable sessions show that review is completed, recurring errors are reducing, delayed retrieval is stable and the next session begins without substantial review debt. Reverse the increase if explanations become superficial, familiar items dominate or recovery deteriorates.

Sources

1. RCPCH - Theory exams: structure and syllabi
2. RCPCH - Changes to theory examinations in 2024
3. Trumble et al. - Systematic review of distributed practice and retrieval practice in health professions education
4. Serra et al. - The use of retrieval practice in the health professions: a state-of-the-art review
5. Castro et al. - The use of feedback in improving the knowledge, attitudes and skills of medical students: a systematic review and meta-analysis
6. Yang et al. - Testing boosts classroom learning: a systematic and meta-analytic review
7. Wendsche et al. - The impact of supplementary short rest breaks on task performance: a meta-analysis