Every week, someone in my coaching sessions asks a version of the same worried question: "How many formulas do I need to memorize for the PMP?" And they're usually holding some ancient list of 40+ formulas from an old forum thread, half of which haven't mattered in years.

So let me give you the honest, current answer — the 2026 exam has fewer pure-calculation questions than the exams of a decade ago, but the ones that appear are the easiest points in the building if you've prepared them, and the concepts behind them show up constantly in situational questions. Here's how I sort every formula into three buckets for my students.

Bucket 1: Memorize cold (the ones that print money)

These come up again and again, both as direct calculations and hiding inside scenario questions:

  • The EVM core four — PV, EV, and the two variances (CV = EV − AC, SV = EV − PV). One memory hook carries you: EV always comes first, and negative is never good.
  • The two indexes — CPI = EV ÷ AC and SPI = EV ÷ PV. Below 1.0 means trouble on that dimension.
  • The default forecast — EAC = BAC ÷ CPI. If the exam doesn't tell you the overrun was a one-time event, this is your formula.
  • Communication channels — n(n−1) ÷ 2. A thirty-second calculation the exam still loves.

Bucket 2: Understand the concept (the formula is secondary)

For these, the exam rarely asks you to compute — it asks whether you know what the number means and what a project manager does about it:

  • TCPI — the honesty metric. When a scenario shows CPI at 0.83 but someone promises to finish on budget, TCPI is how you call the bluff. Understand that TCPI above ~1.10 means "re-baseline, don't hope."
  • VAC and ETC — the sponsor's questions: "how big is the surprise?" and "how much more money do you need?" Know which is which.
  • Float — you should be able to compute LS − ES, but the tested skill is judgment: a delay on a task with float needs monitoring, a delay on the critical path needs action.
  • PERT — (O + 4M + P) ÷ 6 still appears, but more often the exam tests why you'd use a three-point estimate at all: because single-point estimates hide uncertainty.

Bucket 3: Skip without guilt

The old lists are full of formulas the modern exam essentially never tests directly: present value and NPV calculations (understand that higher NPV wins when comparing projects — that's it), depreciation math, point-of-total-assumption contract calculations, and sigma-level quality arithmetic. If your study time is scarce — and whose isn't — spend those hours on situational judgment instead. The return is ten times higher.

How to actually learn Bucket 1 in one evening

Don't make flashcards of naked formulas. Learn each one attached to a tiny story — the same worked example every time. For CPI, mine is always the same project: earned $35k, spent $42k, CPI 0.83, "eighty-three cents of work per dollar." Once the story is in your head, the formula reconstructs itself under exam pressure, even at question 150 when you're tired.

To make that easy, we built a free PMP formula sheet — all 15 formulas that matter for 2026, each with a plain-English meaning and a worked example, on one page. There's a print-ready PDF too, free, no email required. Keep it beside you for one evening of practice and Bucket 1 is done.

Then test it under pressure

Knowing formulas at your desk and retrieving them at minute 190 of a four-hour exam are different skills. After your formula evening, run them inside real exam conditions: our free PMP practice exam mixes calculation items into realistic situational sets, exactly the way the real exam does. When you can hit the calculation questions in under a minute each — because at 80 seconds average per question, that's the pace that banks time for the hard scenarios — the formulas have done their job.

Fifteen formulas. Three buckets. One evening plus one practice session. That's the whole assignment — everything else you've seen on those 40-formula lists is history.