PERT Calculator

PERT calculator for the PMP exam: compute the expected duration and standard deviation from optimistic, most likely, and pessimistic estimates, with a worked example.

By Abu Anns · Last updated 20 November 2026Exam version: July 2026 ECO (current since 9 July 2026)

Three-point estimating and PERT

When a duration (or cost) is uncertain, a single estimate hides the uncertainty. Three-point estimating asks for three numbers instead: the optimistic estimate (O — everything goes well), the most likely estimate (M — the realistic case), and the pessimistic estimate (P — everything goes wrong). The PERT (Program Evaluation and Review Technique) formula combines them into a weighted average that gives the most likely estimate four times the weight of either extreme:

Expected duration = (O + 4M + P) / 6

Standard deviation = (P − O) / 6

The standard deviation tells you how uncertain the estimate is: a wide spread between optimistic and pessimistic means a large standard deviation and a less reliable estimate. The exam also tests the range around the expected value: the expected duration plus or minus one standard deviation covers about 68% of possible outcomes, and plus or minus two standard deviations covers about 95%.

Worked example

An activity has an optimistic estimate of 4 days, a most likely estimate of 7 days, and a pessimistic estimate of 16 days.

So the activity is expected to take 8 days, with about a 68% chance of finishing between 6 and 10 days (8 ± 2), and about a 95% chance of finishing between 4 and 12 days (8 ± 4). Note that the simple average of the three estimates would be 9 days — the PERT formula's weighting of the most likely estimate pulls the answer toward 7, giving 8. The exam tests this distinction: if a question gives you O, M, and P, use the PERT formula, not the simple average.

PERT vs triangular distribution

The exam sometimes contrasts PERT with the triangular distribution, which uses a simple average: expected = (O + M + P) / 3. For our example, the triangular expected duration would be (4 + 7 + 16) / 3 = 9 days, versus the PERT value of 8 days. The difference is the weighting: PERT assumes the most likely estimate is more probable than either extreme, so it weights M four times. If a question gives you three estimates and asks for the PERT expected duration, use the (O + 4M + P) / 6 formula. If it asks for the triangular distribution, use the simple average. The question will name the method — read it carefully, because the two answers differ.

PERT also scales from a single activity to a whole project. The expected duration of a path is the sum of the expected durations of its activities, and the standard deviation of the path is the square root of the sum of the variances (each activity's variance is its standard deviation squared) of the activities on the critical path. That is how a project manager can say, for example, that there is a 95% chance the project finishes within a given date range. The exam rarely asks for the full project-level calculation, but it does expect you to know that path durations add and that uncertainty compounds along the critical path.

When to use PERT

PERT is used when there is significant uncertainty about an activity's duration and the estimator can give a realistic range. It is more work than a single-point estimate, so it is used for critical or high-uncertainty activities rather than for every activity on the project. The exam also contrasts PERT with other estimating techniques: analogous estimating (based on similar past work — fast but rough), parametric estimating (based on a statistical rate, such as cost per square metre), and bottom-up estimating (detailed estimates from the lowest level — the most accurate and the most time-consuming). A question asking for the most accurate technique is asking for bottom-up estimating; a question giving three estimates and asking for the expected value is asking for PERT.

PERT calculator

Enter your three estimates and click Calculate.

Advertisement