The History of Critical Chain Project Management

By Victor Hoang · Updated Oct 21, 2026

By May 1957, Remington Rand and duPont had begun a cooperative effort to test critical path scheduling. Remington Rand supplied programs for duPont's UNIVAC I in Newark, Delaware. That method, the critical path, answered a plain question. Which tasks control the end date?

Critical Chain Project Management (CCPM) was introduced by Eliyahu M. Goldratt in his 1997 business novel Critical Chain. It's the project branch of the Theory of Constraints he had been building since the late 1970s. Forty years after the critical path, Goldratt asked a different question. What controls the end date once people, time padding and shared resources are part of the picture? The throughline is theory first, practice later: an idea from factory floors reached project managers as a story, then spent decades being tested, trimmed and argued over. For the method itself, see what critical chain project management is.

Timeline at a Glance

Before the Chain (1957-1978)

Critical path scheduling leaves people out of the plan

Picture a UNIVAC I in Newark, Delaware, where duPont engineers supplied a small pilot problem to test the method. Kelley and Walker's 1959 paper records that the fundamentals were developed in early 1957 and that by May the theory had advanced far enough for Remington Rand and duPont to start a cooperative effort to implement it. Today that work is taught as the critical path method.

The same paper is candid about what the method left out. It rests mainly on a project's technological requirements, and the authors write that considerations of available manpower and equipment are "conspicuous by their absence." Schedules the technique produces are technologically feasible but not necessarily practical. Their answer was a separate manpower-leveling step: when people aren't available, pick a longer schedule and check the force requirements again.

PERT, the Navy's parallel network method, earned a bigger reputation. The publisher's description of Harvey Sapolsky's study of the Polaris program says PERT gained the program a "worldwide reputation for managerial innovativeness", and our PERT chart guide covers the technique. Both network methods answered "what depends on what?" Another ingredient was behavioral: the habit of using every minute allowed. Herroelen and Leus, who later wrote a rigorous academic review of critical chain, cite Parkinson's Law, that work expands to fill the time allowed, as one of the problems critical chain was built to fight.

Critical chain has no documented ancient ancestor. Its nearest ancestors are these late-1950s network methods, so this history starts there. For the wider arc, see the history of project management.

What this means for teams today: A schedule that is correct on paper can still be unworkable if the same people are needed in three places. The authors of the first critical path paper saw that in 1959.

Goldratt's Factory (1979-1996)

A scheduling program, a novel and a theory

The story moves from project networks to a factory problem. Goldratt was born in 1947 and trained as a physicist, with a bachelor's degree from Tel Aviv University and a master's and doctorate from Bar-Ilan University, according to the TOC Institute. After helping several Israeli manufacturers, he left academia to join a company called Creative Output. PMI's 2011 obituary in PM Network, written by Realization CEO Sanjeev Gupta, says he joined Creative Output in 1979 and that the firm ranked sixth on the Inc. 500 list in 1984. Its product, Optimized Production Technology (OPT), was billed as the first software to provide finite capacity scheduling for production environments. Simons and Simpson's 1997 review notes that Goldratt's production ideas, dating from the late 1970s, travelled under several names: OPT, drum-buffer-rope, synchronized manufacturing and the Theory of Constraints.

Then he wrote a novel. The Goal, with co-author Jeff Cox, came out from North River Press in 1984. The TOC Institute's biography says it was rejected by all major publishers and then published with help from North River Press's owner. By 1987 the Christian Science Monitor reported more than 100,000 copies sold with no advertising budget, bought by over 5,000 corporations, with General Electric among the early buyers of the OPT software. The book is where the Theory of Constraints reached a general audience, and our profile of Goldratt's leadership covers the man and the book at more length.

The decade after was about spreading the factory version. The Avraham Y. Goldratt Institute was established in 1985. The Race, written with Robert Fox, appeared in 1986, and What Is This Thing Called Theory of Constraints in 1990, whose table of contents includes the five steps of focusing. The core logic is the same one behind any business bottleneck: find the constraint, protect it, and don't let other parts of the system pretend to be efficient at its expense.

Projects came later, and first in academic form. A 2012 PMI literature review notes that Ingram and Scherer's 1992 thesis helped blaze the trail by applying the Theory of Constraints to project scheduling. The review records that this groundwork existed before the book. What the 1997 novel did was introduce the project version to practitioners, in the same story format as The Goal.

What this means for teams today: TOC is a thinking tool first and a scheduling tool second. If you can name the one resource or step that limits your output, you've done the first half of what this era taught.

The Novel Arrives (1997-2000)

A business novel carries constraints into projects

The artifact is a 246-page paperback from North River Press in Great Barrington, Massachusetts, with "A business novel" on the cover, according to the library catalog record. It taught through a story, as The Goal had. The 2012 PMI review says the book caused substantial interest among project managers in short order, and that Goldratt himself suggested these ideas were perhaps among the biggest breakthroughs in project management since PERT and CPM. The same year, the TOC Institute's biography says, he retired from the Institute just before his 50th birthday.

The PMI review summarizes the book's argument in a few moves:

• Safety is hidden in every estimate. Goldratt holds that people pad task durations, that each management layer pads again, and that estimators protect themselves against global cuts.
• The padding is then wasted. He cites the student syndrome (starting at the last minute), multitasking, delays where paths merge, and dependencies that pass delays down the line but not early finishes.
• The fix is to aggressively estimate, then pool. Strip the safety from each task, and put the pooled safety from the critical chain into a project buffer that protects the due date. Safety from other paths goes into feeding buffers that protect the chain.
• Control by buffers, not milestones. The chain is the longest sequence of dependent steps, where dependence includes sharing a person. When buffer consumption passes set thresholds, management acts.

Herroelen and Leus add the details a planner would need. Task durations are set at 50% confidence rather than the 80 to 90% common in practice, the project buffer is usually sized at half the project duration, and feeding buffers at half the length of the non-critical path feeding them. They quote Goldratt calling multitasking "the biggest killer of lead time."

The ecosystem formed fast. A 2012 PMI review lists early book reviews in 1998, Robert Newbold's Project Management in the Fast Lane (St. Lucie Press, 1998) and, in 2000, Lawrence Leach's Critical Chain Project Management. Herroelen and Leus list ProChain and Thru-Put Technologies software as available by 1999, and say the Primavera scheduling engine was announced to incorporate critical chain. They also say the principles had been adopted by a growing number of companies.

What this means for teams today: The book's most durable contribution was a diagnosis, not a formula. Padding hides in estimates, and the time it buys gets spent before the work starts.

The Evidence Fight (2000-2010)

Academics test the buffer rules and find limits

If the 1990s were about enthusiasm, the 2000s were about proof. The first serious academic test came from Willy Herroelen and Roel Leus, whose paper appeared in the Journal of Operations Management in 2001. They ran a full factorial experiment on benchmark problems and concluded that the 50% rule for buffer sizing may seriously overestimate the protection needed. They also found that regularly updating the baseline schedule and the critical chain gave the best intermediate estimates and the shortest final duration. In the PMI-hosted version of their paper, they note that critical chain focuses on single-unit renewable resources like people and they argue that multitasking isn't always harmful.

Raz, Barnes and Dvir widened the critique in 2003, and a version later appeared in IEEE Engineering Management Review. The PMI review summarizes their twelve concerns. The table lists several of them, plus one evaluator's view.

Defenders and neutral evaluators answered with simulations. The review reports that Budd and Cooper (2005) found a significant time-saving advantage over the critical path method when critical chain is properly implemented, and Lechler, Ronen and Stohr, in an article titled "new project management paradigm or old wine in new bottles?" in 2005, concluded that critical chain is complex to apply but that many of its ideas can be readily adopted. A 2001 George Mason paper archived by the Defense Technical Information Center captured the stalemate: some organizations that implemented the technique report good progress, others none, and some literature calls it revolutionary while other writers call it a coherent compilation of long-known techniques.

The PMI review makes two quieter points. Publication on critical chain peaked in 2005 and fell after. And in 1997 to 2009 it found only two articles that studied an actual implementation thoroughly, both published a decade after the book. The first, by Srinivasan and colleagues in 2007, covered C-5 aircraft overhaul at Warner Robins Air Logistics Center in Georgia, where project time was set at 50% of the original estimate, half of the cut was added back as a project buffer, 120 feeding paths got their own buffers, and Concerto software identified the chain. The reported benefits were more maintenance workload and revenue, plus better responsiveness. The second, by Bevilacqua and colleagues in 2009, covered a gasification plant at an oil refinery in Falconara Marittima, Italy, and reported shorter task durations and more reliable completion dates against a CPM turnaround project at the same site the year before.

What this means for teams today: Take the diagnosis seriously and the numbers with care. Copy-pasting the "cut estimates in half, add a buffer of half" rule is the part the critics tested hardest.

The Specialist's Tool (2010s-present)

A niche method adopted in constrained, multi-project shops

By the 2010s, the best-documented adopters were organizations with scarce specialists and many projects competing for them. PMI's 2011 obituary for Goldratt, who died on 11 June 2011, says the method is claimed to enable a 10 to 50 percent improvement in project speed and that seven of the eight largest U.S. military air depots reportedly used its techniques. The author is the CEO of a critical chain software company, so treat those figures as vendor-side claims.

A government case with its own documents is NASA's Dryden Flight Research Center. Its paper says that in fiscal year 2011 the center implemented CCPM because audits found multitasking had been "carried to a whole new level," and an individual could work on 10 different projects in a single pay period. The setting matches the method's design: a resource-constrained, multi-project environment.

Meanwhile the profession's standard lost interest. A 2017 Brazilian study of the PMBOK Guide and critical chain found the Guide recommended CCPM only superficially, under Time Management. By 2018 practitioners on PMI's community forum were asking what happened to the critical chain method, because it was gone from the Sixth Edition without explanation. A PMI-community blogger who had tried it argued that its lessons survive: limit multitasking on scarce people, pool uncertainty in visible reserves, and watch milestones more than individual tasks. Mixing methods this way is the subject of the history of hybrid project management.

The research has moved on from proving the method to mixing it. A 2025 systematic review in Applied Sciences covered 62 studies from 2014 to 2025. It found most used modeling and simulation, a growing trend of combining CCPM with Scrum, BIM and Lean Construction, and a remaining gap in empirical studies of engineer-to-order and service organizations.

What this means for teams today: Use it where its conditions hold: scarce specialists, many projects, and managers who will stop starting new work. Elsewhere, borrow the pieces. The triple constraint trade-off gets easier when everyone can see how much schedule protection is left.

Key People

Key Facts

Key Facts: Critical Chain History

• North River Press published Goldratt's Critical Chain in 1997, as a 246-page "business novel."
• Goldratt joined Creative Output in 1979, before The Goal appeared in 1984.
• Critical chain task estimates use 50% confidence instead of the 80 to 90% common in practice, with project buffers usually half the project duration.
• A 2001 benchmark experiment found the 50% buffer rule can seriously overestimate the protection a project needs.
• A PMI-published review found only two thorough implementation studies in the 1997 to 2009 literature.
• A 2025 systematic review covered 62 studies and still found few empirical studies in service organizations.

The Pattern Behind the History

We call this the Export, Test, Narrow pattern. It is our reading of the timeline, not a research result.

1. Export. A rule proven on one kind of work is carried to another: from factory flow to project scheduling.
2. Story first. A novel spread the idea faster than evidence could, with books, software and consultants following within three years.
3. Test. Academics ran experiments on the rules of thumb, buffer sizes above all, and found them too blunt.
4. Narrow. The method settled where its assumptions hold, which is shared scarce resources across many projects.
5. Fragments outlive the package. Single-tasking, visible reserves and buffer tracking stayed in use after the standard dropped the name.

What It Teaches Teams Today

• Look for the constraint before the plan. In critical chain, the scarce person or team sets the pace.
• Ask where padding hides, and what it buys. Estimates carry safety, and a deadline can eat it.
• Limit work in progress for scarce people. The method's strongest claim is about multitasking.
• Pool uncertainty in a reserve that everyone can see, and act on how fast it shrinks.
• Test any ratio before adopting it. The "half the project" buffer was the first rule academics challenged.
• Prefer evidence from a real organization over a clean story. In the 1997 to 2009 literature, thorough implementation studies were rare.

For the wider story, read the evolution of project management methodologies, the project management timeline, and who invented project management.

Critical chain's history is a case study in how a good diagnosis can outlast a contested formula. The method narrowed, the standard moved on, and the question it asked, what really limits this project, is still the right one.