Skip to main content
Category: Laws
Type: Queueing and operations flow law
Origin: John D. C. Little, Operations Research (1961)
Also known as: L = λW; WIP = Throughput × Cycle Time
Quick Answer — Little’s Law says that in a stable system, the long-run average number of items inside equals the average arrival (or throughput) rate times the average time each item spends there: L = λW. In operations language, work-in-process equals throughput times cycle time. To cut lead time without losing output, you usually must reduce WIP.

​
What is Little’s Law?

Little’s Law is the identity that, for a stationary system, average inventory in the system equals average flow rate times average sojourn time.
L = λW — the average number of items in the system equals the average arrival rate times the average time spent in the system.
The relation looks simple because it is conservation, not a gadget. If customers (or jobs, tickets, or bottles) keep entering and leaving at steady average rates, the pile inside is fixed by how fast they arrive and how long each stays. Raise either rate or dwell time, and average occupancy rises in proportion.

​
Little’s Law in 3 Depths

  • Beginner: More arrivals or longer stays mean a bigger queue or backlog on average.
  • Practitioner: Measure any two of WIP, throughput, and cycle time; the third is determined in a stable flow.
  • Advanced: Use the law as a diagnostic identity under stationarity; do not treat WIP as an independent dial that can raise throughput without bound.

​
Origin

In 1961, John D. C. Little published “A Proof for the Queuing Formula: L = λW” in Operations Research (Vol. 9, No. 3, pp. 383–387), then at the Case Institute of Technology. Earlier writers had used the relation without a general proof; Philip M. Morse had even challenged readers to find a counterexample. Little gave broad conditions under which the equality holds for stationary processes. The result became a foundation of queueing theory and operations management. Manufacturing texts such as Hopp and Spearman’s Factory Physics popularized the operations form WIP = TH × CT (work-in-process = throughput × cycle time). Lean and Kanban practice lean on the same identity: control WIP to control lead time when throughput is constrained by demand or capacity.

​
Key Points

Little’s Law is most powerful when you treat it as an accounting identity for flow, not as a slogan to “start more work.”
1

Three linked averages, one identity

Pick consistent units: items in system (L / WIP), items per unit time (λ / throughput), and time per item (W / cycle time). In a stable system, knowing two determines the third.
2

Stability is the fine print

The classic statement assumes a system in steady state with finite means—no permanent explosion of backlog and no undefined averages. During a startup surge or a collapse, short windows can mislead.
3

Lead time falls when WIP falls (at fixed throughput)

If throughput is pinned by demand or a bottleneck, cutting unfinished work is the direct lever for shorter cycle time. Starting more jobs often lengthens waits instead of raising completed output.
4

Throughput has physical ceilings

Raising WIP beyond the level needed to feed the bottleneck mainly adds queueing delay. Pair Little’s Law with capacity and variability thinking—or related limits like Amdahl’s Law for serial stages—so you do not confuse “more started” with “more finished.”

​
Applications

Use Little’s Law wherever work piles up and someone claims that starting more will finish more.

Software and product delivery

Limit open pull requests or WIP columns; if merge rate is ~10 items/week and WIP is 40, expect ~4 weeks average age until WIP drops.

Manufacturing and logistics

Estimate lead time as inventory ÷ ship rate; audit whether high inventory is buying throughput or only buying wait.

Service and healthcare queues

For a clinic seeing 20 patients/hour with average 1.5 hours in clinic, expect about 30 patients present on average—useful for space and staffing stress tests.

Personal workflow

Count open tasks and completions per week; if you finish 5/week but keep 25 open, expect a ~5-week average age unless you stop starting new work.

​
Case Study

A standard operations classroom example makes the arithmetic concrete. Suppose a production line completes 50 units per day (throughput) and holds a roughly steady 200 units of work-in-process across stations. Little’s Law implies average cycle time CT = WIP / TH = 200 / 50 = 4 days. Managers who want a 2-day cycle time at the same throughput must drive WIP toward about 100 units—not “start more jobs.” Lean and Factory Physics teaching use this identity to show why WIP explosions lengthen lead time: with throughput capped by the bottleneck, extra inventory mostly becomes waiting time. The boundary note is important: if the line is still ramping up, or if scrap and rework make “items” inconsistent, you must redefine the system and units before trusting the number.

​
Boundaries and Failure Modes

Little’s Law does not say that any WIP level is achievable with any throughput. Machines, staff, and demand set maximum flow; below Critical WIP you may starve the bottleneck, and far above it you mostly buy delay. It also fails as a casual calculator when the system is not stable—growing backlogs, seasonal spikes, or definitions that mix half-finished and finished items. Averages over mismatched windows produce nonsense. A common misuse is reading L = λW as permission to raise WIP to raise throughput indefinitely. Past the point needed to utilize capacity, more WIP usually raises W instead of λ.

​
Common Misconceptions

Clear use requires separating the identity from capacity strategy and from nearby laws about work expansion.
No. Any stable flow of discrete items—tickets, patients, packets, inventory—obeys the same average relationship when definitions are consistent.
No. With a binding bottleneck, extra starts mainly increase WIP and cycle time; completed output may stay flat.
No. You still need trustworthy measurements of two quantities; the law gives the third and checks consistency, it does not invent data.
These pages help connect flow identity to capacity, incentives, and work expansion.

Parkinson's Law

Work expands to fill the time available—often by growing WIP and stretch.

Diminishing Returns

Extra input (including extra WIP) eventually yields smaller throughput gains.

Goodhart's Law

When a measure becomes a target, it can stop measuring what you care about.

Amdahl's Law

Serial stages cap overall speedup—another structural limit on flow.

Brooks's Law

Adding people late can increase coordination delay, lengthening cycle time.

Stein's Law

Unsustainable WIP growth cannot continue forever; it will stop—often painfully.

​
One-Line Takeaway

If you want shorter lead times at the same output, shrink the pile of unfinished work—Little’s Law says the math will not negotiate.