Home / Lean Production / Little’s Law Calculator
Lean Production and Flow
Little’s Law Calculator (WIP = Throughput x Lead Time, Flow Time and Queue Analysis per ISO 22400)
By Zeeshan Abbas · Reviewed by Rimsha Nadeem Anwar, Six Sigma Black Belt
In short: Little’s Law says WIP = throughput × lead time in any stable system. Enter any two below and this tool solves for the third, in consistent units, for manufacturing, queues, or Agile Kanban, and shows how lead time falls as you cut WIP.
Calculate with Little’s Law
WIP = Throughput × Lead time | Lead time = WIP / Throughput
Lead time
4.0hr
- Work in process (WIP)
- –
- Throughput
- –
- Lead time
- –
- Lead time (other units)
- –
- Relationship
- –
Enter any two values to solve for the third.
Industrial engineering methodology and flow-time workflow
This calculator applies Little’s Law, the single relationship that ties together the three numbers every operation cares about: how much work is in the system (WIP), how fast it comes out (throughput), and how long each unit takes to get through (lead time or flow time). Little’s Law states that in any stable system, average WIP equals average throughput times average lead time. Enter any two and the tool solves for the third in consistent units. Its power is that it is distribution-free: it makes no assumption about the order of processing, the variability, or the routing, so it holds for a machining line, a call-center queue, a hospital ward, or an Agile Kanban board equally.
The data workflow runs end to end. You supply two of the three quantities in matching units, WIP in units, throughput in units per time, and lead time in time, and the tool returns the third along with the derived flow metrics: inventory turns, the implied flow time, and the WIP-to-throughput ratio. The value is not just arithmetic; it is that the law exposes the lever. If lead time is too long, Little’s Law proves it can only be shortened by cutting WIP or raising throughput, and it quantifies exactly how much.
A naive reading treats the three numbers as independent and tries to cut lead time by working harder. Little’s Law shows that is impossible at fixed throughput: lead time and WIP move together. The law also carries a hidden precondition, stability, meaning arrivals equal departures over the averaging window; applied to an unstable or transient system it gives a meaningless number. The sections below make the stability assumption and the variability that lengthens real lead time explicit, so the flow time you plan is the flow time the system actually delivers.
Governing equation: WIP equals throughput times lead time
The model is one equation with three symbols, solvable for any one given the other two.
L = λ × W | WIP = Throughput × Lead time | W = L / λ | λ = L / W
The variables and units:
- L (WIP) = average work-in-process, the number of units (or jobs, customers, tickets) in the system, dimensionless count.
- λ (throughput) = the average completion rate, equal to the arrival rate in a stable system, in units per unit time (units/h, units/day).
- W (lead time) = average time a unit spends in the system from entry to exit, in time. Also called flow time or cycle time in queueing, though it is not the per-station cycle time of a line.
Two rules make the arithmetic honest. First, the units must match: if throughput is units per hour, lead time must be in hours, so WIP comes out as a pure count. Second, all three are time-averages over the same stable window; instantaneous values do not obey the law. A useful derived metric follows directly: inventory turns equals throughput over the period divided by average WIP, so a lower WIP at the same throughput means faster turns and less cash tied up in the flow. There is no Imperial versus SI distinction, only unit consistency between throughput and lead time before solving.
Applicable standards and theoretical frameworks
Little’s Law is a proven theorem, but its use in operations sits inside recognized bodies of work.
| Reference | Scope | Effect on this calculation |
|---|---|---|
| ISO 22400-2 | KPIs for manufacturing operations management | Defines throughput rate and work-in-process inventory, fixing the WIP and throughput quantities the law relates. |
| APICS/ASCM Dictionary | Operations and supply chain terminology | Fixes the definitions of lead time, throughput, WIP, and inventory turns used across planning. |
| Factory Physics (Hopp and Spearman) | Science of manufacturing flow | Places Little’s Law with the critical-WIP and best-case-performance results that bound a line’s throughput and cycle time. |
| Queueing theory (Little, Kingman) | Waiting-line analysis | Provides the proof of the law and the variability, utilization, and time (VUT) relationship that explains why real lead time exceeds the minimum. |
Compliance shapes interpretation. ISO 22400 separates WIP that is flowing from WIP that is blocked or waiting, and only the average over a stable window belongs in the law. Factory Physics adds the critical WIP, the level at which a line reaches its bottleneck rate without adding cycle time; below it the line is starved, above it WIP only inflates lead time. Queueing theory supplies the reason real systems sit above the minimum: variability. State the stable window and the WIP definition, or the number is not defensible.
Key input variables and flow classifications
Three inputs, but each has definitions that must be pinned. WIP can be counted as raw, in-process, or finished, and only the inventory actually inside the system boundary counts. Throughput is the completion rate, which equals demand in a stable system, not the installed capacity. Lead time is total time in system including queueing, distinct from the per-station cycle time. The table gives typical flow relationships by domain; measure your own averages rather than assuming them.
| Domain | WIP unit | Throughput unit | Lead-time unit |
|---|---|---|---|
| Assembly line | Parts in the line | Parts per hour | Minutes to hours |
| Machining job shop | Open work orders | Orders per day | Days to weeks |
| Warehouse or fulfillment | Orders in process | Orders per hour | Minutes to hours |
| Service queue | Customers in system | Customers per hour | Minutes |
| Agile Kanban board | Cards in progress | Cards per week | Days |
Why real lead time exceeds the minimum: variability and stability
Little’s Law holds exactly, but the lead time it reports is an average of a real distribution, and three effects push that average above the floor a naive plan assumes.
Variability inflates queue time (the VUT relationship)
Queueing theory (Kingman’s approximation) shows waiting time rises with the variability of arrivals and process times and, critically, explodes as utilization approaches 100 percent. Two systems with identical throughput and WIP averages can have very different lead-time distributions; the more variable one has a longer tail. Little’s Law gives the average, but variability decides how far individual units stray from it.
The stability precondition
The law requires arrivals to equal departures over the averaging window. During a ramp-up, a demand surge, or a shutdown, WIP is changing and the instantaneous ratio does not equal lead time. Applying the law to a transient period, or over too short a window, yields a number that is arithmetically clean but operationally false. Use a window long enough for the system to be in steady state.
WIP that is not flowing
Blocked, waiting, or defective WIP still counts in the inventory but is not being processed, so it lengthens the measured lead time without adding output. Little’s Law makes this visible: if WIP is high but throughput is flat, lead time balloons, which is the signature of a system carrying excess or stagnant inventory rather than a faster one.
Flow rule: at fixed throughput, lead time is strictly proportional to WIP (W = L / λ). You cannot shorten lead time by pushing more work in; that only raises WIP and lengthens it. The only levers are less WIP or more throughput, and Little’s Law quantifies each.
Minimum WIP versus actual WIP and the excess-inventory penalty
There is a floor on WIP. The minimum WIP that still lets a line run at its throughput is the throughput times the raw process time (the sum of the actual work times with no waiting), the critical WIP of Factory Physics. Below it the line is starved and throughput falls; at it the line hits its bottleneck rate with the shortest possible lead time; above it, every extra unit of WIP adds pure lead time and cash with no throughput gain. Actual WIP almost always sits above the critical level because variability and batching demand a buffer, but the gap between actual and critical WIP is quantified waste. The calculator makes the trade explicit: hold WIP as low as the variability allows to keep lead time short, and use the WIP-to-throughput ratio to see how far above the floor you are running.
Reverse-engineering WIP caps and throughput from a target
The law inverts directly, which is the basis of pull control.
- WIP cap from a target lead time: L_max = λ × W_target. Capping WIP at this level (a CONWIP or Kanban limit) guarantees the average lead time, because lead time cannot exceed WIP divided by throughput.
- Achievable throughput from a WIP limit: λ = L / W, so at a fixed WIP cap, throughput and lead time trade off directly.
- Kanban card count: number of cards = L_max, the WIP cap that enforces the target flow time in a pull system.
- Turns from a lead time: shortening lead time at fixed throughput lowers average WIP proportionally and raises inventory turns by the same factor.
For example, to guarantee a 30 minute (0.5 h) lead time on a line running 120 units/h, the WIP cap is L_max = 120 × 0.5 = 60 units; setting a 60-card Kanban limit holds the average lead time at 30 minutes without scheduling each unit.
Five Little’s Law case studies and worked calculations
Case 1: solve for lead time, baseline line
A line holds an average WIP of 60 units and runs at a throughput of 120 units/h. Lead time W = L / λ = 60 / 120 = 0.5 h = 30 minutes. Each unit spends 30 minutes flowing through, and the inventory turns over 120 / 60 = 2 times per hour.
Case 2: solve for WIP from a target lead time
The same line must guarantee a 20 minute (0.333 h) lead time. Required WIP = λ × W = 120 × 0.333 = 40 units. Cutting WIP from 60 to 40 shortens lead time from 30 to 20 minutes at unchanged throughput, and there is no other way to do it without raising the rate.
Case 3: variability and utilization
Two lines both average 60 WIP and 120 units/h, so both report a 30 minute average lead time. Line A runs at 80 percent utilization with low variability; Line B runs at 97 percent with high variability. By Kingman’s relationship, Line B’s queue time and lead-time tail are far longer for the same average, so promising a firm 35 minute due date is safe on A and risky on B. The average from Little’s Law is identical; the variability decides the service level.
Case 4: Agile Kanban WIP limit
A software team completes 15 cards per week and wants an average cycle time of 4 days (0.8 weeks). WIP limit = λ × W = 15 × 0.8 = 12 cards. Setting a board WIP limit of 12 holds the average delivery time near 4 days; raising the limit to 20 would stretch cycle time to 20 / 15 = 1.33 weeks with no more throughput.
Case 5: reverse calculation, throughput from a WIP cap
A cell is capped at 40 units of WIP by a CONWIP loop and must deliver a 25 minute (0.4167 h) lead time. Required throughput = L / W = 40 / 0.4167 = 96 units/h. If the process can only sustain 90 units/h, the target lead time is infeasible at that WIP cap, so either the cap must rise or the process rate must improve, and the law shows which before any change is made.
Shop-floor implementation and continuous improvement best practices
Cut WIP to cut lead time
At fixed throughput, lead time falls only when WIP falls. Cap WIP with a CONWIP or Kanban limit rather than pushing more work in; releasing more only raises WIP and lengthens lead time.
Measure over a stable window
Little’s Law holds only when arrivals equal departures on average. Measure WIP, throughput, and lead time over a steady window, not during a ramp, surge, or shutdown, or the number will mislead.
Attack variability, not just the average
Two systems with the same average lead time perform very differently if one is more variable. Reduce arrival and process variability and avoid running near 100 percent utilization, where lead time explodes for a small rate gain.
Set Kanban limits from the target flow time
The right number of Kanban cards is throughput times the target lead time. Derive the WIP cap from the flow time you promise, then hold it, rather than sizing the board by intuition.
Boundary conditions, mathematical limits, and model assumptions
Little’s Law is exact for any stable system, but its preconditions are strict. It requires conservation of flow, arrivals equal departures over the averaging window, so it does not apply to a transient or growing system, or to a window too short for steady state. It reports averages only; it says nothing about the distribution, so it cannot by itself promise a due date, which needs the variability and utilization from queueing theory. It counts every unit inside the boundary equally, so defining the system boundary (a station, a line, a whole plant) changes all three quantities and they must be consistent. The law does not care about processing order, so it holds under FIFO or any other discipline, but the individual lead-time spread does depend on discipline. Finally, Little’s Law describes flow, not capacity: it does not tell you the maximum throughput, which is set by the bottleneck, only how WIP and lead time relate at whatever throughput the system actually sustains.
Common Little’s Law mistakes and data interpretation pitfalls
- Confusing lead time with cycle time. Lead time is total time in the system including waiting; cycle time is the per-unit pace of a station. Little’s Law uses lead time, not station cycle time.
- Using capacity as throughput. Throughput is the actual completion rate, equal to demand in a stable system, not the installed capacity. Substituting capacity overstates flow.
- Applying the law to an unstable window. During a ramp or surge, WIP is changing and the ratio does not equal lead time; the law needs steady state.
- Mismatched units. Throughput in units per hour with lead time in minutes gives a WIP off by 60. Keep the time units consistent.
- Reading the average as a guarantee. The law gives the mean lead time; individual units vary with system variability, so a due-date promise needs the queueing view, not just the average.
Integration into MES, ERP, and value stream mapping
Little’s Law is the connective tissue of flow across the planning stack. In an MES it links live WIP counts and throughput to an implied lead time, so a rising WIP trend forecasts a lengthening lead time before due dates slip. In ERP and capacity planning it converts a target customer lead time into a WIP cap and an inventory-turns target, sizing how much stock the flow should carry. In value stream mapping it turns the WIP in each process box into the flow time that dominates the timeline, showing where inventory, not processing, is the delay. In pull systems it sets the Kanban or CONWIP card count directly. Because the same throughput and cycle-time inputs feed the takt, cycle-time, and OEE calculations, a consistent application of Little’s Law keeps inventory policy, scheduling, and lead-time promises working from one relationship.
Little’s Law frequently asked questions
What is Little’s Law in one sentence?
In any stable system, the average work-in-process equals the average throughput multiplied by the average lead time (WIP = throughput x lead time). Given any two, you can solve for the third.
What does “stable system” mean?
Stable means arrivals equal departures on average over the measurement window, so WIP is not systematically growing or shrinking. The law holds only in this steady state, not during a ramp-up, a demand surge, or a shutdown.
Is lead time the same as cycle time?
No. In Little’s Law, lead time (or flow time) is the total time a unit spends in the system including all waiting. Cycle time on a line is the per-unit pace of a station. They are different quantities and are not interchangeable here.
How do I reduce lead time using Little’s Law?
At fixed throughput, lead time is proportional to WIP, so reduce WIP (with a CONWIP or Kanban cap) or raise throughput. Pushing more work into the system only increases WIP and lengthens lead time; it does not speed anything up.
Does Little’s Law work for services and Agile Kanban?
Yes. The law is distribution-free and unit-agnostic, so it applies to customers in a queue, tickets in a service desk, or cards on an Agile board. WIP is cards in progress, throughput is cards completed per week, and lead time is the average delivery time.
Why do two systems with the same average lead time perform differently?
Little’s Law gives the average only. Variability in arrivals and processing, and how close the system runs to 100 percent utilization, decide the spread of individual lead times. The more variable system has a longer tail and a worse on-time performance for the same average.
How does Little’s Law set a Kanban or CONWIP limit?
The WIP cap equals throughput times the target lead time. Set the number of Kanban cards or the CONWIP level to that value, and the average lead time is held at the target without scheduling each unit individually.
Related lean production calculators
Pair Little’s Law with the rest of the toolkit. Return to the Lean Production hub for the full set.
Sources, disclaimer, and editorial transparency
The definition, formula, and applications of Little’s Law used here follow recognized operations-management and lean sources, including John Little’s original 1961 result, the Lean Enterprise Institute, and standard queueing-theory references. Definitions follow ISO 22400 for manufacturing KPIs and the APICS/ASCM Dictionary for flow terminology. This calculator and guide were built by Zeeshan Abbas and technically reviewed by Rimsha Nadeem Anwar, a Six Sigma Black Belt industrial engineer; see our Editorial Policy for how each tool is researched, built, and tested.
Results are accurate estimates for planning and education, not certified engineering advice. Little’s Law assumes a stable system and long-run averages; validate outputs against your own measured data and engineering judgment before changing a process, committing capital, or making staffing decisions. See our full Disclaimer. OpsCalculators.com is operated by MAFHH INTERNATIONAL LTD. Your inputs are processed in your browser and are never stored; see our Privacy Policy.