Two ways of finding the longest sequence in a schedule — one built on task logic and dates, the other on resource contention and protected time buffers.
Both methods answer the same practical question — "which chain of work determines my project's finish date?" — but they get there differently and they protect the answer differently. The Critical Path Method (CPM), developed in the 1950s and the backbone of PMP-style scheduling, finds the longest path through a network of tasks using only logical dependencies and durations, then tracks slack (float) on every other path. Critical Chain Project Management (CCPM), introduced by Eliyahu Goldratt in the 1990s, starts from CPM's network but then re-sequences it around resource availability and strips out padding from individual task estimates, replacing it with pooled buffers placed strategically in the schedule. The two methods can — and often do — identify a different chain of tasks as the one that actually governs the finish date.
Goldratt's core observation, drawn from the Theory of Constraints, was that individually padded task estimates get wasted three ways: Parkinson's Law (work expands to fill the time available), student syndrome (people start late because there's slack), and the fact that early finishes on one task rarely get passed forward to the next task's start — while late finishes always do. CCPM's fix is to cut each task estimate to an aggressive-but-achievable level (often the 50% probability estimate instead of the 90%), remove the resource-contention conflicts that CPM's pure logic ignores, and then protect the whole chain with one shared project buffer at the end (plus smaller feeding buffers where non-critical chains merge into the critical chain). Progress is tracked by watching how much of the buffer has been consumed relative to how much of the critical chain is complete — a buffer burning faster than the chain is progressing is the actual early-warning signal, not any single task running late.
Buffers are the most visible difference, but the deeper change is that critical chain redraws the network around resource availability, not just task dependency logic. CPM will happily show two tasks running in parallel even if they require the same welder, the same test rig, or the same approving engineer — because CPM's definition of the critical path never asks "who is doing this work." CCPM explicitly resolves those resource conflicts first (a step called "resource leveling" against a defined resource calendar), and the resulting longest resource-feasible chain — the critical chain — can therefore be a completely different, and often longer, sequence of tasks than CPM's critical path on the exact same project. Only after that resource-constrained chain is identified does CCPM add the compressed estimates and pooled buffers on top.
Explains how the Critical Path Method (CPM) and Critical Chain Project Management (CCPM) can identify different governing sequences for the same project — CPM from pure task-dependency logic and durations, CCPM from resource-constrained logic plus pooled buffers replacing individually padded estimates.
Both methods produce a highlighted chain of tasks on a schedule and both talk about "the critical path/chain," so it is easy to assume CCPM is just CPM plus a buffer bar at the end. In reality the two methods can select entirely different task sequences as critical, because CPM treats every task as if unlimited resources were available to execute it, while CCPM explicitly checks whether the people, equipment, or approvals a task needs are shared with another task running at the same time.
Critical Path Method (CPM): build a network of tasks connected by logical dependencies (finish-to-start, start-to-start, etc.), estimate each task duration (typically already containing some individual safety margin), then forward-pass and backward-pass the network to compute early/late start and finish dates for every task. The path with zero total float — no slack anywhere along it — is the critical path, and it equals the shortest possible project duration given the stated logic and durations.
Critical Chain Project Management (CCPM): start from the same network, then resource-level it — resolve any case where two tasks on different paths need the same constrained resource at the same time, forcing one to wait. Strip the safety margin out of individual task estimates (commonly cutting to roughly a 50% confidence duration) and relocate that removed time into a single project buffer at the project's end, plus smaller feeding buffers wherever a non-critical path merges into the critical chain. The longest resource-feasible sequence, inclusive of these dependencies, is the critical chain.
CPM remains the default scheduling technique taught in PMP and PRINCE2 curricula and built into virtually every scheduling tool (MS Project, Primavera P6) — it is simple, well understood by every stakeholder, and sufficient when resources genuinely aren't the constraint. CCPM earns its added complexity specifically in resource-constrained environments — engineering teams sharing a small pool of specialists, a single test lab, or a limited set of certified inspectors — where CPM's "unlimited resources" assumption silently hides schedule risk that only shows up once two tasks fight over the same person or machine. Program and portfolio-level PMOs sometimes run CCPM specifically to combat the chronic multitasking and late-task-start behavior that individually padded CPM estimates tend to produce across an organization.
Often, yes — even though individual task estimates are cut, and even after accounting for resource leveling that can lengthen the chain, the pooled-buffer approach frequently yields a shorter overall committed date than a traditional CPM schedule loaded with individually padded, unmanaged safety margin in every task.
Not strictly, but mainstream tools like MS Project and Primavera P6 do not natively compute or display buffer consumption the way CCPM requires, so most organizations running CCPM use a dedicated plug-in or a specialized tool (e.g., ProChain) alongside their main scheduling software.
CPM is core, tested material in the PMP exam and forms the backbone of PRINCE2 planning theory. CCPM is generally mentioned only briefly as an alternative scheduling technique — the PMP exam expects you to recognize what it is and how it differs from CPM, not to compute a full CCPM schedule.
No. Float (or slack) in CPM is unused time that already exists in the network because a task can start later without delaying the project — it is a property of the schedule's logic. A CCPM buffer is deliberately inserted, sized time that was removed from task estimates elsewhere and relocated to protect the finish date — it is a managed resource that gets consumed and tracked, not a passive byproduct of scheduling logic.
Try our Engineering Management, Business & Professional Skills Studio
More calculators, simulators, and guides for this discipline.
Data Analysis Studio
✨ Premium ContentA structured, paid professional training program for this discipline.