Every earned value figure on the PMP exam comes from three numbers — planned value, earned value and actual cost — combined in two ways. Subtract for a variance, divide for an index, and earned value always comes first. Learn that shape and the whole set rebuilds from memory in under a minute.
Most candidates meet the earned value formulas as a list of about a dozen equations, memorize them the week before the exam, and then confuse two of them under time pressure. The list is the problem. Learned as a list, the formulas have no internal logic, so there is nothing to fall back on when memory goes.
They have a shape. Every earned value figure on the PMP exam is built from three numbers, combined in two ways. Once you can see that, you can rebuild the whole set from memory in under a minute — which is exactly what you want to be able to do on a scratch pad before the first question.
Planned value is what you said you would have accomplished by now, in money. Earned value is what you have actually accomplished, valued at the plan's own prices. Actual cost is what you spent getting there. PV, EV, AC. Every other figure comes from these.
They combine in exactly two ways. Subtract, and you get a variance — a gap expressed in currency. Divide, and you get an index — a ratio where 1.0 means on plan.
So: cost variance is EV minus AC. Schedule variance is EV minus PV. Cost performance index is EV divided by AC. Schedule performance index is EV divided by PV.
One rule fixes half the errors people make on exam day: earned value always comes first. It is the first term in every subtraction and the numerator in every division. If you have written AC minus EV, you have written it backwards.
A negative variance is bad news, and an index below 1.0 is the same news said differently. Positive and above 1.0 are good. That is the whole interpretation.
Which one is about cost and which about schedule follows from the second term. Actual cost is money you spent, so anything paired with AC is a cost statement. Planned value is the schedule's promise, so anything paired with PV is a schedule statement.
Work one example and it stops being abstract. A project budgeted at 100,000 is six months in. The plan said 50,000 of work would be done; 40,000 has actually been earned; 45,000 has been spent. Cost variance is minus 5,000, schedule variance is minus 10,000, CPI is 0.89 and SPI is 0.80. The project is behind and over — and the SPI being worse than the CPI says the schedule is the more urgent problem.
Estimate at completion is where candidates lose points, because there are four of them and they look interchangeable. They are not. Each answers a different assumption about the future, and the scenario always tells you which assumption to make.
If the variances you have seen so far are typical of what is coming, EAC is BAC divided by CPI. This is the default when the scenario gives you nothing else.
If the overrun was a one-off — a one-time expedite fee, a supplier problem now resolved — the rest of the work runs at plan, and EAC is AC plus the remaining budgeted work, BAC minus EV.
If both cost and schedule performance will keep dragging on the rest of the work, EAC is AC plus the remaining work divided by CPI multiplied by SPI.
And if the original estimate is simply no longer credible, you re-estimate the remaining work from the bottom up: EAC is AC plus a new ETC. Two more fall out of these: estimate to complete is EAC minus AC, and variance at completion is BAC minus EAC.
The exam question is almost never "which formula". It is a paragraph that tells you something about the future, and the formula is your reading of that paragraph. Read for the assumption first.
To-complete performance index is the efficiency you would have to achieve from here on to still land on a target. Skipping it is a mistake, because it is the formula that turns numbers into a decision.
Against the original budget, TCPI is the work remaining over the money remaining: BAC minus EV, divided by BAC minus AC. Against a revised target, when the original budget is already gone, the denominator becomes EAC minus AC.
The reason it matters is the comparison. If your CPI has been running at 0.85 and TCPI comes out at 1.40, the arithmetic is telling you the team would have to become far more efficient than it has ever been. That is not a budget you recover; that is a conversation with the sponsor. Exam scenarios use this gap deliberately.
The exam that began on 9 July 2026 tests judgment in context, and earned value is no exception: the number is rarely the answer by itself. You compute it in order to decide what to do next, and the answer options are usually actions rather than figures.
So practice in that direction. Compute the index, say out loud what it means, then choose the response — rather than stopping at the arithmetic. Our visual study guide draws these relationships as decision maps rather than listing them, which is the format that survives a long exam: fifty maps across People, Process and Business Environment, with the earned value set drawn as one shape you can redraw from memory.
As always, PMI's current published instructions on test-day rules and materials win over anything a study guide tells you.
Plan on reproducing these yourself rather than counting on anything being provided, and check PMI's current test-day rules for what you may use. The practical answer is that a set you can rebuild from three numbers is worth more than a sheet you have to find.
PMI publishes percentages at the domain level only — People 33 percent, Process 41 percent, Business Environment 26 percent on the front of the July 2026 outline — and none for individual topics. We will not invent a number. What is safe to say is that the formulas are cheap to learn properly and expensive to guess at.
BAC divided by CPI. It is the assumption that what you have seen is what you will keep seeing, and it is the default the exam expects when nothing points elsewhere.
They pair: the main guide explains, the visual guide shows you the shape of the decision. The fifty maps also stand on their own as a revision layer in the final weeks, when re-reading chapters is no longer the best use of your time.
Rules change. Where a figure or a procedure can move, the issuing agency’s current published instructions win over anything here.
