How to Answer PMP Situational Questions: The Decision Framework That Works on the New Exam

The new PMP exam launched today, July 9, 2026. We covered the July 2026 exam change in detail before launch, but the short version: domain weights moved from 42/50/8 to People 33%, Process 41%, and Business Environment 26%, new interactive item types arrived alongside multiple choice, and the content outline now emphasizes value delivery, outcomes, strategic alignment, adaptive and hybrid delivery, AI in projects, sustainability, and stakeholder engagement. Still 180 questions in 240 minutes.

What did not change: the exam is dominated by situational questions, and situational questions are where prepared candidates lose points. This post gives you the answer-selection framework, updated for the new outline, with three worked examples.

Why situational questions feel unfair#

Read enough PMP practice questions and you will hit the same frustration: all four options are things a reasonable project manager might actually do. Nothing is factually wrong.

That is by design. The exam is not testing which option is true. It is testing which action comes first, or best fits the situation, among several defensible ones. The stem almost always contains the words that matter: "what should the project manager do FIRST," "do NEXT," "what is the BEST course of action."

Candidates who study PMP like a knowledge exam fail it as a judgment exam. You cannot memorize your way past 180 scenarios. You can internalize the decision style PMI rewards, because it is remarkably consistent.

The PMI mindset, in eight principles#

Every correct situational answer traces back to a small set of values. Learn the why behind each one, because the exam varies the surface details endlessly while the underlying logic repeats.

1. Address problems directly before escalating. Escalation hands your problem to someone with less context; PMI expects the PM to own it first.
2. Never choose "do nothing" or "wait." A PM who postpones a known problem is accepting unmanaged risk, and the exam treats passivity as the worst action available.
3. Engage stakeholders early and directly. Almost every project failure mode gets cheaper to fix the earlier the affected people are in the conversation.
4. Find root cause before acting. Acting on symptoms produces rework, so options that start with "assess," "understand," or "meet with" usually beat options that start with "implement" or "instruct."
5. Protect value and outcomes over process for its own sake. Process exists to deliver benefits, and an answer that follows procedure while destroying value is a trap.
6. Empower and support the team rather than command. PMI's model PM is a servant leader, so options where the PM overrides, decides for, or bypasses the team are almost always wrong.
7. Follow change control for scope, cost, or schedule changes. Uncontrolled change is how baselines die, so any change to an approved baseline routes through impact assessment and a change request, even in hybrid projects.
8. Tie decisions to business value and strategy. This one matters far more on the new exam, because the Business Environment domain roughly tripled in weight (8% to 26%), and the outline explicitly centers value delivery and strategic alignment.

If you failed a previous attempt on scenario judgment, this list is probably why. Our post on why people fail the PMP on the second attempt goes deeper on how to turn missed questions into a pattern log instead of repeating the same instincts.

The four-step elimination framework#

Here is the framework to run on every situational question. With practice it takes under a minute.

Step 1: Identify what is really being tested, and the moment in time. Find the FIRST, NEXT, or BEST in the stem. Has the problem just been discovered, or has analysis already happened? An answer that is correct at minute 30 is wrong at minute 1.

Step 2: Strike options that skip investigation, skip people, or jump to escalation. Cross out anything that acts before understanding, excludes the affected stakeholders or team, or sends the problem up the chain as the opening move.

Step 3: Strike options that violate change control or abdicate responsibility. Cross out anything that changes a baseline informally ("just add it to the sprint"), refuses engagement ("that is not my job"), or parks the problem ("log it and revisit later").

Step 4: Between the finalists, pick the answer closest to the root cause and highest in stakeholder and value alignment. Usually two options survive steps 2 and 3. The winner is the one that understands before it acts and that keeps business outcomes, not just compliance with process, in view.

Now let's run it on three exam-realistic questions, one per domain.

Worked example 1: People (33%)#

Two senior engineers on your hybrid team disagreed in the daily standup about the design of an upcoming integration. The disagreement has continued for a week, it is becoming personal, and two other team members have started staying silent in meetings. What should the project manager do FIRST?

A. Escalate the conflict to the engineers' functional managers so it can be resolved formally B. Select one of the two designs yourself so the team can move forward C. Meet with the two engineers privately to understand the source of the disagreement D. Reassign one of the engineers so the integration work is no longer blocked

Step 1. This is a conflict-management question, and the moment is FIRST: nothing has been investigated yet.

Step 2. A escalates as the opening move; strike it. B and D both act without investigating, so they are suspect, but hold them for a moment.

Step 3. D abdicates the PM's conflict-resolution responsibility by removing a person instead of resolving the problem; strike it.

Step 4. Between B and C: B is the trap, and it is attractive because the schedule pressure is real and "the team can move forward" sounds like protecting value. But B is command behavior that treats the symptom (no decision) rather than the root cause (an unresolved technical and interpersonal conflict), and it teaches the team that disagreements get settled by authority. C investigates root cause, engages the people directly, and keeps the PM in servant-leader mode. Answer: C.

Worked example 2: Process (41%)#

A key stakeholder in a hybrid project asks you to add a reporting feature to a component being delivered under the project's predictive scope baseline. The feature looks small, and the development team confirms they have spare capacity in the current sprint. What should the project manager do NEXT?

A. Ask the team to include the feature in the current sprint since capacity exists B. Decline the request and explain that the scope baseline was approved and cannot change C. Assess the impact of the request on scope, schedule, cost, and expected value, then submit a change request D. Escalate the request to the sponsor for a decision

Step 1. This tests change management in a hybrid context, at the NEXT step after a request lands.

Step 2. D escalates before any analysis exists for the sponsor to decide on; strike it.

Step 3. A changes an approved baseline informally; the spare capacity is bait, because capacity does not make an uncontrolled scope change controlled. Strike it. Note the hybrid wrinkle: if this feature lived purely in the adaptive backlog, routing it to the product owner for prioritization could be right. The stem says it hits the predictive scope baseline, which is exactly why re-reading the stem matters.

Step 4. Between B and C: B is the trap for rule-followers. It sounds disciplined, but it is process for its own sake, a "do nothing" wearing a governance costume. Baselines are not frozen; they are changed through change control. C assesses impact (including value, which the new exam cares about) and routes the decision properly. Answer: C.

Worked example 3: Business Environment (26%)#

Your project is delivering a customer data platform scheduled to go live in five months. A regulator announces a new data-residency rule that takes effect one month before your go-live date. What should the project manager do FIRST?

A. Add the regulation to the risk register and review it at the next monthly steering committee meeting B. Direct the team to pause development until the compliance department confirms the requirements C. Analyze the regulation's impact on the project's deliverables, compliance obligations, and expected business value, and bring options through change control D. Escalate to the sponsor and request additional budget for compliance rework

Step 1. This is an external business environment change, at FIRST discovery.

Step 2. D escalates and asks for money before anyone knows what the rework even is; strike it.

Step 3. B pauses delivery, which is "wait" dressed up as prudence, and it also hands the analysis entirely to another department. Strike it.

Step 4. Between A and C: A is the trap, and it is genuinely attractive because logging risks and using governance cadences is textbook behavior. But a regulation with a known effective date before go-live is not an uncertain risk to monitor; it is a change to the project's compliance obligations that threatens the platform's ability to deliver value at all. Waiting for a monthly meeting violates the "never wait" principle. C analyzes impact in value terms and uses change control. Answer: C.

Expect a lot more questions shaped like this third one. At 26% of the exam, Business Environment is now roughly 47 of your 180 questions.

The new item types change your tactics, not your mindset#

The framework works on every format, but the new interactive items reward specific habits:

• Case-study question sets share one scenario across several questions. Re-read the stem for each item, because the moment in time usually advances between questions, and the FIRST action for question two is often the thing question one just did.
• Drag-and-drop and matching items reward knowing sequences and mappings cold: process orders, conflict-resolution approaches, which artifact serves which purpose. You cannot eliminate your way through a matching item; you either know the pairs or you do not.
• Point-and-click and graphic interpretation items put the data in a chart or diagram. Read the axes and labels before the question, the same way you would read a stem.
• Budget your time. 240 minutes for 180 questions is 80 seconds per question. Interactive items take longer than multiple choice, so bank quick wins and flag anything that stalls you past two minutes.

How to practice until the mindset is automatic#

Volume of scenario questions plus reading every explanation, including on the ones you got right, because guessing correctly for the wrong reason is a hidden gap. Start untimed and run the four steps deliberately, writing down which principle eliminated each option. Once your accuracy stabilizes, switch to timed sets at the 80-second pace, then full-length timed exams for stamina.

If you keep missing the same shapes (escalating early, informal scope changes, "wait" answers), that pattern is fixable and worth more than any additional reading. And if you are still deciding whether this cert is worth the effort at all, start with who should take the PMP before you sink the hours.

Ready to drill the mindset until it is automatic? Start a free practice test on ExamCoachAI. The free tier gives you 10 questions a day, enough to run one framework rep every morning before work.