Critical Path Analysis in Jira

The critical path is the longest chain of dependent tasks through your project — the chain with zero float, where any slip pushes the end date. Native Jira doesn't compute it. Structure.Gantt, BigPicture, and Foundation's Gantt all do. Here's what the algorithm actually means and how to use it in planning conversations.

What is critical path analysis (and what isn't it)?

The critical path method (CPM) was developed in the late 1950s by DuPont and Remington Rand to schedule chemical plant shutdowns, and converged with PERT (from the US Navy's Polaris program) into the standard scheduling model behind every modern Jira Gantt app.

The critical path is the longest sequence of dependent tasks from project start to finish. Tasks on it have zero float. A one-day slip on any critical task is a one-day project slip. Tasks off the critical path have positive float and can slip up to that amount without moving the deadline.

What it isn't: a list of the most important tasks, the highest-priority work, or whatever's blocked right now. It's a structural property of the dependency network plus task durations. A nice-to-have can sit on the path; a business-critical task can be off it if it has parallel slack.

What's the exact algorithm in plain language?

CPM runs two passes. The forward pass walks start to finish, computing Early Start (ES) and Early Finish (EF) for each task. The backward pass walks finish to start, computing Late Start (LS) and Late Finish (LF). Total float is LS − ES. Tasks with zero float are critical; the chain of those tasks from start to end is the critical path.

Worked example. Four tasks:

A: Design — 5 days, no predecessors
B: Build backend — 10 days, depends on A
C: Build frontend — 6 days, depends on A
D: Integration test — 3 days, depends on B and C

Forward pass: A finishes day 5. B runs day 5–15. C runs day 5–11. D waits for both, so D runs day 15–18. Project duration: 18 days. Backward pass: floats come out A = 0, B = 0, C = 4, D = 0.

Critical path: A → B → D. Task C has four days of float — it can slip four days without moving the deadline. The insight: criticality is about position in the dependency network, not task weight.

Why doesn't Jira Plans compute critical path?

Jira Cloud's native Plans view (formerly Advanced Roadmaps, bundled with Jira Premium) shows dependency icons when an issue is blocked, but stops there. It doesn't run forward/backward passes, calculate float per issue, or highlight the longest chain. Atlassian positions Plans as a portfolio roadmap, not a scheduler.

Practical effect: if a program manager asks “which tasks are on the critical path?”, native Jira can't answer. You need a Marketplace app that implements CPM on top of Jira data. See the pillar guide on Jira Gantt charts for the broader context.

Which Jira Gantt apps compute it and how do they display it?

Three Marketplace apps compute classic CPM on Jira data. Structure.Gantt toggles CPM from a toolbar button; critical bars render with a red outline and a float column appears in the grid. BigPicture Gantt includes CPM in its Scheduling module with red highlighting and adds optional resource-leveled critical path. Foundation toggles CPM from the Gantt toolbar; critical bars render red with float available as a column in any Lens view. Foundation maps dependencies directly to Jira issue links, so CPM runs on the same “blocks” graph visible elsewhere in Jira.

What's float, and how is it different from buffer?

These terms get used interchangeably in casual talk but they're mathematically different.

Float is calculated by CPM — it falls out of the forward/backward pass. It's a property of the network, not a decision.

Buffer is deliberately added as a contingency reserve, usually at the end of a chain or project. The PMI PMBOK Guide calls this “management reserve” or “contingency reserve.” Goldratt's Critical Chain Method replaces per-task float with explicit project and feeding buffers.

In practice: float is what the tool tells you; buffer is what you add because you know the math assumes perfect estimates. Healthy plans have both.

How should you actually USE the critical path in a status meeting?

Knowing which tasks are critical only matters if it changes the meeting. Three patterns work well:

Front-load critical tasks in the walkthrough. Start with the critical chain, not the top of the backlog. A one-day slip on a critical task is a one-day project slip; the same slip on a task with five days of float is a non-event.

Ask “why is this critical?” on surprises. When an unexpected task appears on the path, a dependency has changed — someone added a blocker, a predecessor grew, or a parallel chain shortened. The CPM result is a diagnostic signal.

Staff by criticality, not by squeaky wheel. Per PMI scheduling guidance, critical-path tasks deserve disproportionate oversight and priority access to shared resources. Tasks with weeks of float rarely need daily senior attention.

What are near-critical paths and why should you track them?

A near-critical path is a chain with small but positive float — typically one or two days in a multi-week project. These chains are one surprise away from becoming critical, and often do. A program reporting “critical path is green” while four near-critical chains quietly consume their float is in worse shape than it looks.

Good Gantt tools let you view the top N longest paths, not just the strict critical one. Structure.Gantt offers a configurable threshold; Foundation exposes float as a sortable column so you can filter under a given slack budget. The baseline guide covers when to snapshot after replans; for dependency mechanics, see Jira Gantt charts with dependencies. Critical path math is only as good as the graph underneath it.

Frequently asked questions

What exactly is float (slack) in critical path analysis?

Float is the amount of time a task can slip without extending the project end date. Total float is calculated as Late Start minus Early Start (or equivalently Late Finish minus Early Finish). Tasks on the critical path have zero total float by definition — any delay moves the deadline. Tasks with positive float have slack to absorb, but consuming that slack converts a non-critical path into a critical one, which is why near-critical tracking matters.

Does Jira Premium compute critical path?

No. Jira Premium includes Plans (formerly Advanced Roadmaps) which shows dependency icons when an issue is blocked, but it does not compute the critical path, does not calculate float, and does not visualize the longest chain. If you need critical path inside Jira Cloud, you need a Marketplace app such as Structure.Gantt, BigPicture, or Foundation.

How often should the critical path be recomputed?

Whenever the plan changes materially. For fast-moving projects that usually means weekly; for agile teams running a Gantt alongside Scrum, at sprint boundaries works well. Critical path tools recompute automatically on every schedule edit, so the more useful cadence question is how often you formally review it in a meeting. Per PMI guidance, a weekly critical-path walkthrough is standard for mid-sized programs.

Can a project have more than one critical path?

Yes. When two or more chains tie for the longest duration through the network, all of them are critical. This is more common than it sounds — any project with parallel workstreams that share a deadline can produce multiple critical paths. Large programs often also have near-critical paths (chains with only one or two days of float) that deserve tracking because a small slip promotes them to critical.

What is the difference between float and buffer?

Float is calculated by the scheduling algorithm — it falls out of the network math based on task durations and dependencies. Buffer is deliberately added by the planner as a contingency reserve, usually at the end of a path or project. Critical Chain Method (Goldratt) replaces task-level float with explicit buffers. In a standard CPM Gantt you are working with float; if you want buffer, you add it as tasks or milestones.

Does critical path account for resource constraints?

Classic CPM does not — it assumes unlimited resources and computes the longest dependency chain. If a single person is scheduled on two parallel tasks, standard critical path will not flag it. Resource-leveled critical path (sometimes called critical chain) does account for resource constraints and can produce a different answer. BigPicture supports resource leveling; Structure.Gantt and Foundation compute classic CPM and expose resource views separately.