The 2026 PMP exam is mostly judgment, not math. But the calculation questions that do appear are the cheapest points on the test: the numbers compute cleanly, an on-screen calculator is provided, and there's exactly one right answer — no mindset ambiguity. This is every formula worth knowing beyond earned value, each with a worked example. For CV, SV, CPI, SPI, EAC, ETC, VAC, and TCPI, use our dedicated EVM formulas guide — we won't repeat it here.
Float and the critical path
Total float is how long an activity can slip without delaying the project:
Total float = LS − ES = LF − EF
Free float is how long it can slip without delaying its earliest successor:
Free float = ES (of earliest successor) − EF (of the activity)
You find ES/EF with a forward pass (left to right, taking the latest predecessor finish) and LS/LF with a backward pass (right to left, taking the earliest successor start).
Worked example. Activity A (3 days) is followed by B (4 days) and C (2 days) in parallel; both feed D (2 days). Path A–B–D takes 3 + 4 + 2 = 9 days; path A–C–D takes 3 + 2 + 2 = 7 days. The critical path is A–B–D at 9 days, where every activity has zero float. C's forward pass gives ES = 3, EF = 5; its backward pass gives LF = 7, LS = 5. Total float of C = 5 − 3 = 2 days — and its free float is also 2, since D's ES is 7. On the exam, sketch the network on your whiteboard; don't do passes in your head.
75 free questions, scored against the new domain weights. No card required.
Three-point estimates (PERT)
Two averaging schemes, built from Optimistic, Most likely, and Pessimistic values:
- Beta (PERT): E = (O + 4M + P) / 6
- Triangular: E = (O + M + P) / 3
- Standard deviation: SD = (P − O) / 6
Worked example. O = 4 days, M = 7, P = 16. Beta: (4 + 28 + 16) / 6 = 48 / 6 = 8 days. Triangular: 27 / 3 = 9 days. SD = (16 − 4) / 6 = 2 days, so a one-SD range is 6 to 10 days. If the stem says "PERT" or "beta," weight the most likely value; if it says "triangular" or "simple average," don't. The distractors will always include the other scheme's answer.
Communication channels
Channels = n(n − 1) / 2, where n is the number of people.
Worked example. A team of 8 has 8 × 7 / 2 = 28 channels. Two people join: 10 × 9 / 2 = 45, so the question "how many new channels?" is 45 − 28 = 17 — the classic trap is answering 45. The second trap is n itself: "a team of nine plus the project manager" means n = 10. Count every communicator in the stem before you square anything.
Expected monetary value (EMV)
EMV = probability × impact, with threats negative and opportunities positive. For a decision tree, multiply along each branch and sum.
Worked example. A project faces a 30% chance of a $40,000 weather delay (EMV = −$12,000) and a 20% chance of a $25,000 early-completion bonus (EMV = +$5,000). Net EMV = −$7,000, which is the logic behind sizing a contingency reserve. Exam stems usually stop at one multiplication — the skill being tested is keeping the sign and the probabilities straight.
Project selection: ROI, NPV, and friends
The exam rarely makes you compute these by hand; it makes you choose with them.
- ROI = (benefit − cost) / cost. A project costing $100,000 that returns $125,000 has ROI = 25,000 / 100,000 = 25%.
- NPV: future cash flows discounted to today, minus cost. Bigger is better — and NPV already accounts for the time value of money, so when Project A has an NPV of $150,000 over five years and Project B has $175,000 over three, you pick B without adjusting for the durations. That adjustment is the trap.
- BCR (benefit-cost ratio): above 1 means benefits exceed costs; higher wins.
- IRR: the discount rate at which NPV hits zero; higher wins.
- Payback period: time to recoup the investment; shorter wins, but it ignores everything after payback — which is why NPV outranks it when a stem offers both.
Point of total assumption (PTA)
For fixed-price incentive fee contracts:
PTA = (ceiling price − target price) / buyer's share ratio + target cost
Worked example. Ceiling price $110,000, target price $100,000, target cost $90,000, share ratio 80/20. PTA = (10,000 / 0.8) + 90,000 = $102,500 — the cost point past which the seller absorbs every additional dollar. This one appears rarely; learn the shape of it and move on.
What not to memorize
ITTO counts, earned-schedule variants, depreciation arithmetic, and control-chart trivia beyond "three standard deviations" are poor investments for 2026. The list above plus EVM covers what the exam actually pays for. Spend the reclaimed hours on situational practice.
Making formulas stick until exam day
Formulas decay faster than concepts — that's the forgetting curve doing its normal work, and rereading a cheat sheet doesn't stop it. Retrieval with spacing does: close the sheet, produce the formula from memory, apply it, and repeat at growing intervals. That's exactly how the test-retest method treats calculation questions in ExamElephant — miss a float or EMV item and it returns at 3 and 7 days until you've mastered it, so the formula is recallable, not just recognizable. Try it on the free 75-question tier and see which formulas survive a week. It also explains why you won't need a giant brain dump sheet at the testing center, and where formula drills fit inside a 90-day study plan.
Paid passes are coming soon
The full 2,100-question bank, timed 180-question mock exams and unlimited tutoring. Leave your email and we'll tell you the day they open. Unsubscribe anytime. Privacy
FAQ
Do I get a calculator on the 2026 PMP exam?
Yes — an on-screen calculator is built into the testing interface, at Pearson VUE centers and online. You cannot bring a physical one. The arithmetic is deliberately clean; the test is whether you know which formula applies.
How many calculation questions should I expect?
PMI doesn't publish a count, but calculation items are a small minority of the 180 questions. They're still worth preparing: they're fast, certain points that subsidize time for the long situational stems.
Beta or triangular — which do I use if the stem doesn't say?
"PERT" alone implies the beta version, (O + 4M + P) / 6. Stems that want the triangular version nearly always say "triangular" or "simple average." When in doubt, beta — and check whether the distractors include both results.
Do I need to discount cash flows to calculate NPV myself?
No. The exam gives you NPV figures and tests whether you pick the highest one — including when a longer time period tempts you to "adjust" a number that's already time-adjusted.