PWPM Wiki

Program Evaluation and Review Technique

PERT (Program Evaluation and Review Technique) estimates activity durations probabilistically using three points — optimistic, most likely and pessimistic — to produce a weighted expected duration and a measure of uncertainty.

PERT is a scheduling and estimation technique that handles uncertainty by asking, for each activity, three questions: how long if everything goes well (optimistic, O), how long normally (most likely, M), and how long if things go badly (pessimistic, P). It combines them into a weighted average — the expected time — that leans toward the most likely value but accounts for the tails. PERT was developed for the US Navy’s Polaris programme, where activity durations were genuinely unknown, and it pairs naturally with the critical path method.

Program Evaluation and Review Technique at a glance

Category
Methodologies & Frameworks · Scheduling & Time Management · Tools & Techniques
Type
Methodology / framework
Also known as
PERT, PERT Analysis
Appears in
3 sections
Related
Critical Path Method, Theory of Constraints, Critical Chain Project Management

Also known as: PERT, PERT Analysis.

Why it matters

Single-point estimates ("this takes 6 days") hide the uncertainty that actually drives project risk. PERT makes that uncertainty explicit and quantifiable: the standard deviation tells you how confident to be, and summing variances along the critical path lets you state a completion date with a confidence level (for example, an 80% chance of finishing by a given date) rather than a single optimistic number that is almost always wrong.

When to use it

Use PERT when activity durations are uncertain and the cost of being wrong is high — research, new-technology, first-of-a-kind work. It adds effort (three estimates instead of one) so it is overkill for routine, well-understood tasks. Many teams apply it selectively to the riskiest activities rather than the whole plan.

How to use it

  1. For each activity, gather three estimates: Optimistic (O), Most Likely (M) and Pessimistic (P).
  2. Compute the expected duration: TE = (O + 4M + P) ÷ 6.
  3. Compute the standard deviation: SD = (P − O) ÷ 6, and variance = SD².
  4. Use TE values to find the critical path.
  5. Sum the variances of critical activities, take the square root for the path’s SD, and use it to state completion confidence.

Example

An activity is estimated at O = 4, M = 6, P = 14 days. TE = (4 + 24 + 14) ÷ 6 = 7 days. SD = (14 − 4) ÷ 6 ≈ 1.67 days. So you would plan 7 days but recognise a realistic spread of roughly 3.7 to 10.3 days at two standard deviations.

Template

A PERT worksheet captures O, M and P per activity and auto-calculates TE, SD and variance — see the calculator below.

Browse templates →

Tools

Excel / Google SheetsMicrosoft ProjectPrimavera P6@RISK / Monte Carlo add-ins

Formula & calculator

TE = (O + 4M + P) ÷ 6 · SD = (P − O) ÷ 6 · Variance = SD²

Try it yourself: PERT Calculator computes this from your own figures.

Open the PERT Calculator →

FAQs

Why is the most likely estimate weighted by 4?
PERT assumes a beta distribution where the most likely value dominates but the extremes still pull the average. The 4M weighting produces a mean that is close to reality for skewed estimates.
What is the difference between PERT and CPM?
CPM uses single, deterministic durations; PERT uses three-point probabilistic durations. In practice they are combined: PERT for the estimates, CPM for the network and critical path.
Is PERT the same as three-point estimation?
PERT is a specific weighted form of three-point estimation. A simple triangular three-point estimate averages the three values equally; PERT weights the most likely value four times.

Alternatives

  • Critical Path Method — deterministic durations
  • Monte Carlo simulation — models the full distribution, not just three points
  • Story-point / velocity estimation — the Agile alternative