CPA REG Exam Format: Questions, Testlets and Timing

At a glance

These counts come from the published blueprint. Some questions may be pretest items that are evaluated but do not contribute to a candidate’s score. Candidates are not told which items are pretest, so approach each as if it counts.

The five testlets

REG begins with two MCQ testlets followed by three TBS testlets. Each MCQ testlet contains 36 questions. Simulation testlets contain two, three and three tasks. This means preparation needs to cover both formats: efficient MCQ work leaves time for simulations, but rushing through a testlet can create avoidable errors.

Candidates can review items within a testlet under the exam software controls. After submitting a testlet and moving on, they cannot return to it. Use the review screen before continuing and reserve time to check unanswered or flagged questions.

Multiple-choice questions

MCQs test knowledge and application across the REG blueprint. A question may ask for a tax result, a next step or the legal effect of an event. Read the complete stem, identify the taxpayer, entity, tax year and transaction facts, then determine which rule the question tests.

For calculations, write a short sequence rather than jumping to an answer. This can expose missing basis adjustments, filing-status assumptions or excluded amounts. Estimate where practical and use options to catch arithmetic or classification mistakes, not to replace reasoning.

Task-based simulations

A TBS presents a case with exhibits or other information and asks you to complete one or more tasks. REG simulations may require calculations, information selection, analysis of records or a written response. They test how you use facts together, not just whether you recognize a definition.

Open exhibits and identify their date, source, taxpayer and relevance. Not every fact is useful. Record the requested deliverable, units and tax year before calculating. If a task has several fields, solve each separately and check that the figures are consistent.

Do not spend unlimited time on one difficult field. Complete parts you can solve, flag uncertain pieces and move on. Partial progress is preferable to leaving later tasks untouched.

Build a four-hour pacing plan

The 240 minutes cover all five testlets. One starting allocation might reserve about 75 minutes for MCQs and 165 minutes for simulations, then adjust after timed practice reveals your actual pace. This is a study framework, not an official AICPA allocation. TBSs vary in complexity.

1. Track time after each testlet in mock exams.
2. Set a soft target for MCQs and move on when a question consumes too much time.
3. Scan simulation exhibits before calculating.
4. Complete solvable simulation fields before returning to difficult ones.
5. Review each testlet before submission because you cannot return later.

How REG is scored

AICPA reports a scaled score from 0 to 99, with 75 passing. The score is not the percentage of questions answered correctly. For REG, MCQs and TBSs each contribute 50 percent. Do not infer a fixed raw number of correct answers from the scaled score.

Provider practice percentages and readiness estimates can help monitor your own study, but they are not official REG scores. Use errors and blueprint coverage to decide what to review.

Practise the actual interface

Use the official AICPA sample test to explore the software and response formats. Then practise with a timer using materials aligned to the current blueprint. A sample test teaches navigation; it does not estimate readiness. Include MCQs and simulations because both count materially.

More practice and interpretation

Pacing should be trained with checkpoints rather than a fixed number of minutes per question. A candidate may answer a straightforward tax item quickly and need longer for a simulation with several exhibits. After each testlet, compare the remaining time with your plan. If you are behind, choose the next action deliberately: finish a calculation you understand, make a reasoned selection and move on, or flag a specific question for a later check. Do not reread every earlier item simply because the clock feels uncomfortable. For a simulation, a scratch outline can be brief: requirement, taxpayer, tax year, relevant inputs, calculation, final requested measure. For example, if a prompt asks for taxable income and gives wages, interest, an adjustment, and a deduction, write the stages before entering a total. This catches a common slip in which a candidate gives adjusted gross income because that was the last number calculated. For an entity task, add a separate line for basis and limitations instead of folding them into a single number. Practice transitions between question types. Complete a short multiple-choice set, then a simulation, then return to a flagged item. This tests whether you can reset attention after a different kind of work. In review, note the time spent as well as correctness. If a wrong answer took four minutes, the remedy may be a decision rule for moving on rather than another content lecture. If a quick wrong answer came from a missed word such as “except,” practice marking the requested output and qualifier before calculating.

Using the information in practice

A complete timed rehearsal should include the full mix of question types and a post-exam review. During the sitting, note only broad checkpoints, such as whether the second half is starting on schedule. Afterward, compare planned time with actual time by task. If your MCQ pace is slow because you repeatedly change a supported answer, practice committing once you can explain the choice. If simulations take too long because you read every exhibit twice, practice reading the requirement first and annotating relevant evidence on the initial pass. The answer interface also affects execution. Practice entering responses in the same general format as the task: selecting a category, filling a numeric cell, or writing a concise explanation. Before leaving a simulation, scan for an unanswered field and confirm the response addresses the requested item. Do not use the last minutes to rewrite a clear answer from scratch. A short final check for period, taxpayer, sign, and units can catch a preventable error.

Additional application and review

One timed practice set should deliberately test the move from short questions to longer tasks. For example, complete a mixed MCQ block, pause only for the planned transition, and then work through simulations in order of confidence and estimated effort. If a simulation appears lengthy, read the requirement first. Some exhibits may be unnecessary, and identifying the deliverable can show which ones matter. Avoid spending the entire remaining time perfecting one response while other fields remain blank. For review, distinguish a pacing error from a content error. If you knew the rule but spent four minutes toggling between two options, practice a decision rule: eliminate unsupported choices, select the best remaining answer, and flag it only if a specific fact can be checked later. If you missed a rule, timed repetition alone will not fix it. Return to the concept, state it plainly, and solve a fresh item. Track both accuracy and time so the next practice cycle addresses the actual problem.

A final practical check

Keep a short review list during practice that pairs each flag with a reason. “Check whether the prompt asks for recognized or realized gain” is actionable. “Hard question” is not. Once you return, verify the named distinction and either correct the answer or keep your original choice. This limits circular reconsideration and protects time. After several timed sets, compare how many flags changed your answer correctly. If changes are usually wrong, spend less review time changing answers and more time identifying decisive facts on the first pass.

Putting the method into a scenario

A final five-minute review is useful only when the response can be checked. Scan for blanks, ensure a numeric answer belongs in the requested field, and confirm that the unit and sign match the prompt. Do not reopen every completed multiple-choice item. If a flagged answer has no new evidence, keep the reasoned selection. Build this habit in practice so the review period is not spent replacing answers from anxiety.

Additional decision example

If your practice results show a mismatch between accuracy and time, change one variable at a time. First try a set with the same length but mixed topics. If timing improves, the earlier slowdown may have come from topic switching. If it does not, set a soft limit for first-pass review and record how often a later check changes the answer. This creates evidence for your pacing plan instead of relying on a single difficult practice session.