Timed or untimed MRCPCH questions: when to switch

Use both untimed and timed MRCPCH question practice, but give them different jobs. Start important new or unstable material without a strict countdown so you can build a repeatable question process and complete feedback. Introduce soft time boundaries once that process is reliable, then use full timed blocks to rehearse pacing, flagging and recovery. Switch according to error patterns and completion data, not a universal accuracy percentage or a fixed number of weeks before the exam.

This guide covers non-clinical examination preparation for Foundation of Practice (FOP), Theory and Science (TAS) and Applied Knowledge in Practice (AKP). It contains no clinical teaching, diagnosis, prescribing, treatment, safeguarding or patient-management guidance.

The direct answer: train reasoning first, then train it under constraint

Untimed and timed practice answer different questions.

• Untimed practice asks: Can I understand the task, choose a defensible answer and learn from the feedback without the clock distorting the process?
• Timed practice asks: Can I execute a stable process, make proportionate decisions and complete the planned block within a real constraint?

Neither mode is automatically more rigorous. Untimed work becomes weak when it turns into open-book searching or unlimited rumination. Timed work becomes weak when candidates rush through unstable reasoning, accumulate unread feedback and mistake speed for mastery.

The timer should test a process you have built; it should not replace the process you still need to learn.

The practical solution is a staged transition: learning mode, bridge mode and rehearsal mode.

Key points

• Use untimed but bounded practice to develop a complete attempt-review-repair loop; untimed should not mean unlimited, distracted or open-book by default.
• Add timing progressively after the question process is stable, measuring accuracy, completion, confidence, flagging and review quality rather than one headline percentage.
• Keep regular full timed blocks before the examination, but return selected error patterns to untimed repair when time pressure repeatedly damages reasoning.

What the current MRCPCH timing means for practice

The RCPCH structure page, last updated on 27 August 2026, states that the three MRCPCH theory examinations are computer-based. Foundation of Practice (FOP) and Theory and Science (TAS) each contain 100 single best answer questions and last two hours. Applied Knowledge in Practice (AKP) contains 120 single best answer questions split into two papers of 60, with 150 minutes for each paper.

Those totals imply broad average rates:

• FOP and TAS: 120 minutes for 100 questions;
• AKP: 150 minutes for each 60-question paper.

They do not mean every question deserves exactly the same number of seconds. Question complexity, reading load and confidence vary. A useful pacing plan protects the total paper while allowing deliberate variation between items.

RCPCH explained that its 2024 changes to FOP and TAS timing followed analysis of candidate time use and a common SBA timing convention. It also retained more time for AKP because its analysis found that candidates spent longer on the more complex questions. That is evidence about examination design, not proof that every learning session should copy live timing from the first day.

For the current official structure, always check the RCPCH structure and syllabi page. The PaedVault exam time-management guide converts those totals into an exam-day pacing and checkpoint plan. This article has a different purpose: deciding how much time pressure to use while learning.

Why untimed practice is not the same as easy practice

Good untimed practice removes a performance constraint while preserving retrieval.

You still:

1. read the question without notes;
2. identify the precise task;
3. generate a decision or comparison rule;
4. commit to one answer;
5. record confidence;
6. check the explanation; and
7. repair the reason for an important error.

The difference is that the clock does not force you to abandon a reasoning step before you understand it.

Untimed practice is particularly useful when:

• the topic or question style is new;
• you cannot yet explain why one option is preferable;
• correct answers depend on guesses or vague familiarity;
• the same reasoning error keeps returning;
• you are learning a new review method; or
• a timed score falls because the underlying process is inconsistent, not merely slow.

The word bounded matters. If one ordinary question consumes 20 minutes because you repeatedly search notes, the session is no longer a clean retrieval attempt. Set a generous boundary, commit the best answer available, then use feedback to learn.

The MRCPCH question-bank study method explains the full attempt-review-retest loop. The self-explanation guide shows how to examine reasoning without writing an essay after every question.

Why timed practice is a separate skill

The live examination requires decisions under a fixed total duration. Candidates therefore need practice in:

• moving on when extra thought is unlikely to change the answer;
• recognising when a question deserves a flag;
• protecting later questions from one early time sink;
• maintaining a consistent reading process under pressure;
• recovering after an uncertain item; and
• using checkpoints without checking the clock after every sentence.

These behaviours cannot be fully learned in unlimited conditions. Timed blocks reveal whether a method survives constraint.

But a timer does not diagnose the cause of every mistake. A wrong timed answer might reflect:

• missing knowledge;
• an unstable decision rule;
• a misread task;
• premature commitment;
• excessive checking;
• a poor flagging rule;
• loss of concentration; or
• a genuinely difficult item that should have been parked.

Timed practice becomes useful only when the review separates these causes.

What learning research can and cannot tell us

Practice testing supports learning, but does not prescribe one timer

A systematic and meta-analytic review integrated 222 classroom studies with 48,478 students. Practice testing improved learning overall compared with several alternative study conditions. Effects varied with test format, corrective feedback, repetition, setting, timing and study design.

The review supports doing questions and checking feedback. It does not identify an MRCPCH-specific timed-to-untimed ratio or transition score.

Feedback must remain part of the session

A systematic review of feedback in medical students included 26 studies, with 13 included in its meta-analysis. Feedback was associated with better educational outcomes than control conditions, although the authors reported high heterogeneity.

The practical implication is modest: do not raise question volume or tighten the timer so far that explanations and repairs become permanent backlog. A completed timed block without completed feedback is only half a learning loop.

The feedback-timing guide compares reviewing after each item with reviewing after a question block.

Time pressure can change several outcomes at once

An experiment with 211 adults learning unfamiliar arithmetic facts compared timed and untimed learning across different spacing schedules. Time pressure increased perceived stress. Its effect on later speed depended on the spacing condition, while untimed practice showed advantages for some unpractised and conceptually related items. The authors explicitly warned that results from one material type may not generalise to others.

Another experimental study compared timed and untimed learning of picture pairs in children, adolescents and adults. Untimed conditions produced higher recall, while slight time pressure sometimes increased efficiency but weakened aspects of metacognitive self-regulation.

Neither task resembles an MRCPCH paper. Together they show why simple claims such as "timed is always better because it is realistic" or "untimed is always better because it builds accuracy" overstate the evidence. Time pressure can alter stress, stopping rules, reflection, speed and accuracy in different directions.

The honest conclusion

There is no direct evidence that one universal sequence guarantees better MRCPCH performance. The defensible approach is to use both conditions, preserve feedback and measure whether reasoning transfers as constraints increase.

Use SHIFT to move from untimed to timed practice

SHIFT is a PaedVault editorial framework, not an RCPCH requirement:

1. S - State the purpose of the block
2. H - Hold a stable question process
3. I - Introduce a soft boundary
4. F - Full timed rehearsal
5. T - Track what the timer changes

S - State the purpose of the block

Label every question block before it starts.

Mode	Primary purpose	Clock condition
Learning	Build reasoning and feedback habits	Untimed but bounded
Bridge	Preserve reasoning while reducing excess time	Visible elapsed time or generous block limit
Rehearsal	Practise live-paper execution	Current exam-realistic total timing
Repair	Rebuild one recurring failure	Untimed or lightly timed

This label prevents score confusion. A 75% learning block completed with generous pauses does not mean the same thing as a 75% first-seen timed block. Both may be useful; they answer different questions.

Record the mode in PaedVault notes or your revision log alongside the date, question source, first-seen status and review completion.

H - Hold a stable question process

Before tightening time, check whether you can repeat a short process:

1. identify the task;
2. find the discriminating information;
3. compare the plausible options;
4. commit with a stated reason; and
5. move to feedback or the next item.

You do not need perfect accuracy. You do need enough consistency to tell the difference between a knowledge gap and a clock-created process failure.

Signs that the process is not yet stable include:

• rereading the stem without knowing what decision is required;
• selecting an option before comparing alternatives;
• changing answers because time feels uncomfortable rather than because evidence changed;
• being unable to explain correct answers; and
• leaving most feedback unread.

Keep those patterns in learning or repair mode. Use the best-of-five strategy guide to simplify the underlying decision process before adding pressure.

I - Introduce a soft boundary

Do not jump directly from unlimited practice to a full examination.

Bridge methods include:

• run a timer upward to observe actual time without a countdown;
• use a generous total block limit rather than a strict per-question alarm;
• practise five to ten questions with one checkpoint at the end;
• allow one planned extension, then review why it was needed; or
• tag questions that exceed the boundary without forcing a random answer.

The goal is information: which part of the process consumes time?

For example, a candidate may discover that reading is quick but answer changes consume four minutes. Another may answer rapidly but spend too long opening explanations during the attempt. Those problems require different repairs.

Avoid setting a timer for every individual question. Frequent alarms can fragment attention and teach candidates to react to the clock rather than manage the paper as a whole.

F - Full timed rehearsal

Once the process survives bridge blocks, schedule current exam-realistic practice.

A full timed block should rehearse:

• the correct total time for the selected paper or block;
• first-seen questions where possible;
• no notes or external searches;
• a predeclared flagging rule;
• planned checkpoints;
• a brief reset after difficult items; and
• complete post-block feedback.

Full-paper practice is not required every day. It is cognitively expensive and produces a large review burden. Use smaller timed blocks for frequent execution practice and occasional larger simulations for stamina and pacing.

The mock-exam strategy explains how to schedule larger rehearsals. The official sample-paper guide shows how to preserve the limited first-seen value of RCPCH specimens.

T - Track what the timer changes

Compare like with like. For each block, record:

• first-seen or repeated questions;
• learning, bridge, rehearsal or repair mode;
• attempted and completed items;
• accuracy;
• low-confidence correct answers;
• unanswered or rushed items;
• time at each planned checkpoint;
• recurring error causes; and
• whether feedback was completed.

Then compare a topic or mixed block under two conditions.

Pattern	Likely interpretation	Next action
Similar accuracy, complete within time	Process transfers under constraint	Continue mixed timed practice
Accuracy falls and misreads rise	Constraint is disrupting execution	Use shorter bridge blocks and repair reading process
Accuracy stable but many unanswered	Decisions are too slow	Practise stopping and flagging rules
Fast completion with avoidable errors	Timer is causing premature commitment	Restore a mandatory comparison step
Timed block completed but feedback abandoned	Volume exceeds learning capacity	Reduce block size and protect review time
Repeat scores improve but unseen scores do not	Familiarity may be inflating performance	Use fresh questions for verification

The question-bank score guide explains why percentages require context. The daily question-target guide helps set a volume that still permits full review.

When should you switch to timed MRCPCH practice?

There is no validated universal threshold such as 70%, 80% or four weeks before the exam.

Move from learning mode to bridge mode when, across several sessions:

• you use the same concise question process;
• most correct answers are justified rather than lucky;
• repeated errors are being repaired;
• feedback is completed within the session plan; and
• excess time comes from identifiable behaviours rather than basic confusion.

Move from bridge mode to full timing when:

• the process remains intact under a generous boundary;
• checkpoints no longer cause frantic changes;
• you can park a difficult item deliberately;
• accuracy and confidence do not collapse merely because the clock is visible; and
• you have enough time after the block to analyse errors.

These are transition signals, not pass criteria.

When should you return to untimed repair?

Timed practice should generate targeted repair, not become a one-way progression.

Return a pattern to untimed or lightly timed work when:

• the same misread happens in several timed blocks;
• you cannot explain why your chosen option was preferable;
• high-confidence errors increase;
• flags become a way to avoid decisions;
• you finish only by guessing the final section; or
• review shows that the underlying rule was never secure.

Repair the smallest unit possible. Do not convert one difficult timed block into a complete restart of every topic.

The weekly revision review can be used to decide which patterns stay timed and which return temporarily to repair mode.

A four-week example transition

This is an example framework, not an official schedule or a universal prescription.

Week 1: observe the process

• Use mainly untimed but bounded first-seen sets.
• Record elapsed time without enforcing a countdown.
• Identify the two largest sources of delay.
• Complete feedback before adding more volume.

Week 2: add bridge blocks

• Introduce short mixed sets with a generous block limit.
• Use one checkpoint, not per-question alarms.
• Compare accuracy, confidence and completion with learning-mode sets.
• Repair one timing-related behaviour at a time.

Week 3: rehearse current timing

• Add several exam-paced blocks using fresh questions.
• Practise a clear flagging and stopping rule.
• Keep selected unstable patterns in untimed repair.
• Review every rushed, unanswered and low-confidence item.

Week 4: integrate

• Use one larger timed simulation where appropriate.
• Protect sleep and normal routine rather than adding daily full mocks.
• Keep short repair blocks for repeat failure patterns.
• Finalise checkpoints and the post-question reset routine.

If your examination is further away, stretch the phases. If it is closer, do not skip reasoning repair entirely; use smaller cycles within the available days.

Common mistakes

Doing every question timed from the beginning

This may rehearse rushing before a reliable process exists. Speed on unstable reasoning is not the same as efficient performance.

Staying untimed until the final week

Knowing how to solve an item without constraint does not show that the method survives a fixed paper. Leave enough time to practise execution, flagging and recovery.

Using a per-question countdown as the main pacing tool

Questions vary. Manage the total block with checkpoints and a stopping rule instead of expecting identical timing for every item.

Counting unfinished feedback as completed practice

Question attempts generate data; feedback turns that data into learning. Reduce volume if review cannot be completed.

Comparing repeated untimed questions with unseen timed questions

Exposure and timing both changed, so the percentages are not comparable. Label question history and conditions.

Interpreting a slower careful session as regression

New material, deeper feedback or a changed question mix can increase time. Look at the process and error types before concluding that performance has worsened.

How PaedVault can support the transition

Use PaedVault to create small, clearly labelled sessions for each mode:

• a focused learning set for a weak area;
• a mixed bridge set with an external total-time boundary;
• a timed first-seen block for execution practice; and
• a selective repair set for incorrect, guessed or low-confidence items.

Keep question history intact so repeated exposure is visible. Record timing condition and purpose in your session notes. Review explanations completely, then verify repairs with different questions rather than immediately repeating the same answer.

Session size should follow the time available for both answering and feedback. The revision-session length guide helps match block size to attention and rota constraints.

Final answer

MRCPCH preparation should not choose permanently between timed and untimed questions. Use untimed but bounded work to build and repair reasoning, bridge blocks to find where time is lost, and full timed blocks to rehearse the actual performance constraint.

The switch should occur when your question process is repeatable and feedback remains complete, not when you reach an arbitrary percentage. Track what the timer changes, move specific failures back into repair mode and keep regular exam-paced practice until pacing and reasoning can coexist.

Frequently asked questions

Should I do MRCPCH questions timed or untimed?

Use both. Start new or unstable material untimed but bounded so you can practise a complete reasoning and feedback process. Add soft time boundaries when that process becomes consistent, then use current exam-paced blocks for pacing, flagging and recovery. Keep selected weak patterns untimed even while the rest of your practice becomes timed.

When should I start timed MRCPCH question practice?

Start bridge timing when you can use a repeatable question process, explain most correct answers, complete feedback and identify where excess time is going. Move to full timed blocks when that process remains intact under a generous boundary. There is no validated universal score or number of weeks that applies to every candidate.

Does untimed MRCPCH practice make me too slow?

Not if it is bounded and followed by a planned transition. Untimed practice becomes unhelpful when it allows unlimited rumination, frequent note searching or no commitment to an answer. Record elapsed time, build a concise process and introduce progressively tighter block limits rather than staying unlimited until the examination.

Should every MRCPCH question have its own timer?

Usually, manage a block with checkpoints rather than using an alarm for every item. Questions vary in reading and reasoning load, so equal per-question limits can encourage unnecessary rushing. Use a stopping rule for individual time sinks while protecting the total paper time defined by the current RCPCH structure.

What should I do if my score drops during timed practice?

Review the cause before adding more timed volume. Separate missing knowledge, misreading, premature commitment, excessive checking, poor flagging and unanswered items. Repair the recurring pattern untimed or with a softer boundary, then retest it on fresh timed questions. A temporary score drop identifies a transfer problem; it does not by itself prove lack of readiness.

Sources

1. RCPCH - Theory exams: structure and syllabi
2. RCPCH - Changes to theory examinations in 2024
3. Testing (quizzing) boosts classroom learning: a systematic and meta-analytic review
4. The use of feedback in improving the knowledge, attitudes and skills of medical students: a systematic review and meta-analysis
5. Distributed practice and time pressure interact to affect learning and retention of arithmetic facts
6. The effect of time pressure on metacognitive control: developmental changes in self-regulation and efficiency during learning
7. Systematic review of distributed practice and retrieval practice in health professions education