PMI-SP critical path questions are usually tests of judgment under changing conditions. The schedule may show a familiar activity name, but the real issue is often elsewhere: an update may be unreliable, a relationship may no longer reflect the work, a constraint may be shaping float, or a resource decision may have changed the path that controls completion.
A strong scheduler reads the question as a forecast investigation. The goal is not to defend the baseline label or select the most forceful intervention. It is to explain what the current model says, identify the evidence that supports it, and choose a response that fits the situation and the available authority.
Read the current network as a forecast
The baseline is useful for variance measurement, but it is not a permanent description of project criticality. Once actual dates, remaining durations, calendars, dependencies, constraints, or resource availability change, the current network may tell a different story.
Start by separating four kinds of information in the scenario:
- Observed performance: actual starts, actual finishes, installed quantities, completed milestones, or verified field conditions.
- Forecast information: remaining duration, expected finish, production rate, or estimated recovery.
- Model assumptions: relationships, calendars, lag, constraints, resource availability, and scheduling rules.
- Proposed decisions: resequencing, added resources, fast tracking, crashing, or a change to a committed date.
This classification prevents a common error in PMI-SP exam questions: treating a proposed response as though it were already supported by the schedule evidence.
Critical, near-critical, and constraint-driven paths
In CPM, the critical path is the longest logical route to the selected completion point under the current schedule assumptions. Its activities commonly have zero or very little total float. A near-critical path is close enough to the driving duration that a modest variance can make it controlling.
For example, three routes lead to equipment handover:
| Route | Forecast length | Schedule meaning |
|---|---|---|
| Survey → design → approval → installation → handover | 42 days | Current longest route |
| Survey → design → testing → handover | 40 days | Two days of tolerance; sensitive to testing progress |
| Survey → procurement → delivery → setup → handover | 36 days | More apparent margin, subject to resources and dates |
If testing requires three additional days, its route becomes 43 days and takes control of the forecast. The activity itself has not become “critical by identity.” Its path has become the longest under the revised conditions.
Float is model-dependent
Total float describes movement available before the relevant completion date is delayed, subject to the model. Free float concerns movement without delaying the early start of an immediate successor. These measures should not be treated as interchangeable.
Negative float usually indicates that the forecast conflicts with a required date. It may be caused by a contractual milestone, regulatory obligation, external delivery date, or an unsupported restriction. Negative float is therefore a signal to investigate the relationship between the schedule and the commitment; it is not an automatic instruction to crash a particular activity.
Five PMI-SP critical path scenarios
1. A verified delay affects current driving work
A foundation pour was expected to finish on Day 18. Field records now indicate completion on Day 22, and the successor cannot begin until the pour is complete.
The useful question is not simply whether the activity is late. Examine the remaining quantity, actual production, work calendar, relationship type, and successor dates. If the update is credible, the forecast should reflect the new work information. The schedule analyst then determines whether the four-day movement consumes float or moves the project finish.
Relevant artifacts include the status record, quantity report, remaining-duration estimate, current network, milestone forecast, variance analysis, and issue log. A recovery recommendation should explain the cause and the forecast consequence.
Changing the activity finish merely to preserve the previous milestone conceals the variance. Adding labor immediately may also be weak if the cause is a design clarification, access restriction, or incorrect production assumption. The better exam answer is the one that establishes the impact before recommending recovery.
2. Progress updates expose a new driving path
At the previous update, procurement was controlling completion. The latest status shows that deliveries are ahead of expectation, while a testing route retains more remaining work than previously forecast. After the current status is reflected in the network, testing becomes the longest route.
This change may represent normal schedule behavior rather than a defective model. Faster performance on one route can reveal the risk carried by another. Inspect actual dates, remaining durations, relationship logic, float trends, milestone forecasts, and the critical- and near-critical-path reports.
Earned value measures can help indicate a performance trend, but SPI does not identify the relationship that drives the finish. That question belongs to schedule network analysis. The practical implication is a change in surveillance: the testing route now deserves closer control, even if procurement remains important for other reasons.
A tempting but weak answer is to continue managing only the original baseline path. Another is to assume that any change in criticality proves the schedule is wrong. Both choices ignore the current forecast.
3. A constraint produces misleading float
A regulatory inspection must occur by a fixed date. Several activities show negative float, although the unconstrained network appears capable of finishing before the overall project completion date.
Identify the source and authority of the date. Is it a legal requirement, a contract commitment, an approved planning assumption, or a manually entered restriction? Then inspect calendars, relationships, late dates, and the constraint register. The displayed float may be measuring distance from the imposed inspection date rather than flexibility against the unconstrained project finish.
If the obligation is valid, retain it and analyze the activities that drive the inspection milestone. If the date is unsupported, route it for correction or approval. Removing a legitimate restriction to make the report look healthier damages schedule credibility; accepting an arbitrary restriction without review creates artificial pressure.
The relevant answer is usually about understanding the model and the commitment, not immediately compressing every activity with negative float.
4. Resource leveling changes the driving path
Two work fronts are logically parallel, but both require the same specialist. With only one qualified person available, leveling postpones one front. The delayed work consumes its float and extends the forecast finish.
Compare the logic-only network with the resource-constrained version. Determine whether the movement came from a resource calendar, assignment limit, leveling priority, explicit dependency, or manual date adjustment. Review resource-loaded reports, assignments, availability, leveling results, and before-and-after milestone dates.
Possible responses include a second qualified specialist, a different sequence, revised work windows, or acceptance of the delay. The best choice depends on technical suitability, safety, cost, productivity, and actual availability. Resource smoothing generally attempts to remain within available float, whereas leveling can move work beyond that tolerance and alter completion.
Adding people automatically is not a complete scheduling strategy. Ignoring the delay because the activity was not critical in the logic-only network is equally weak. The exam is testing whether you recognize that resource decisions can change the effective driving path.
5. A stakeholder requests an earlier finish
A sponsor asks for delivery four weeks earlier because a market event has moved forward. The request is urgent, but urgency alone does not identify a feasible compression plan.
Translate the request into a measurable finish target and examine the activities that currently control that date. Crashing may shorten selected work through additional resources, overtime, or increased expenditure. Fast tracking may overlap sequential work, but it can create coordination problems, rework, technical risk, and quality exposure.
Use a what-if model to compare candidate options. Examine time saved, cost, resource capacity, dependency logic, safety, quality, risk, and the possibility that another path becomes critical after compression. Crashing non-driving work may add cost without improving the finish. Moving the milestone alone creates negative float rather than recovering time.
A stakeholder request is not the same as authorization to change the approved schedule or baseline. Present the alternatives and trade-offs through the relevant governance route before implementing the selected option.
Progress measurement can change the diagnosis
Percentage complete is meaningful only when its basis is clear. A work package reported as 70% complete might reflect installed quantities, milestone rules, elapsed time, or a subjective estimate. Those methods can imply very different remaining durations.
Test the forecast against physical evidence such as production rates, completed quantities, field records, and the difficulty of the unfinished work. An activity can appear far advanced while retaining its most complex portion. Conversely, a conservative update may overstate the remaining effort.
Keep performance measurement separate from network diagnosis. Planned value, earned value, actual cost, and SPI can reveal trends. CPM, float-path analysis, and milestone forecasts show which activities and relationships affect completion. Strong PMI-SP exam answers use these tools together without substituting one for another.
How to select the best answer
When several options seem reasonable, compare them against the evidence and the decision stage represented in the question.
| Question clue | What it usually tests | Weak shortcut to reject |
|---|---|---|
| Field update conflicts with the forecast | Reliability of actuals and remaining work | Intervene before understanding the data |
| Another route becomes longest | Dynamic criticality and path surveillance | Manage only the baseline path |
| Negative float follows a required date | Constraint authority and late-date logic | Remove the constraint cosmetically |
| Leveling postpones parallel work | Resource-constrained schedule analysis | Assume logic-only CPM is sufficient |
| Leadership wants an earlier milestone | Compression feasibility and governance | Edit the milestone without modeling work |
In a critical path scenario, favor the answer that preserves the integrity of the forecast, distinguishes facts from proposals, accounts for near-critical exposure, and matches the action to the authority available. That reasoning is more reliable than memorizing which activity was labeled critical in an earlier update.
