How to add think time in JMeter

Realistic user pacing in JMeter comes from pauses between business actions, not only from the number of virtual users. Adding think time keeps a thread from immediately starting the next request after a response returns, which makes a web or API workload closer to how people browse, read, search, and submit forms.

A JMeter timer under a Thread Group, controller, or sampler affects the samplers in that scope. A Uniform Random Timer is a common fit for think time because it combines a minimum delay with a random extra delay, so virtual users do not all pause for exactly the same number of milliseconds.

Think-time timers slow sampler starts; they do not hold server responses open or enforce an aggregate request rate. Use a throughput timer when the requirement is a fixed request rate, and keep setup, login, or cleanup samplers outside the think-time scope when those actions should not be paced like normal user actions.

1. Open the test plan in the JMeter GUI and select the Thread Group or controller that contains the requests to pace.
2. Add the timer from Add → Timer → Uniform Random Timer.
3. Name the timer for the user action it represents.

   Name: Think time - checkout pause

4. Enter the variable pause window.

   Random Delay Maximum: 500
   Constant Delay Offset: 1000

   JMeter adds the random delay to the offset. These values make each affected thread wait from about 1000 to 1500 milliseconds before the next sampler starts.

5. Keep the timer under only the requests that should receive the pause.

   Timers are processed before each sampler in their scope. Add the timer as a child of one sampler when only that sampler should wait, or place it under a controller when every child sampler should use the same think-time rule.

6. Save the test plan with the timer in place.

   think-time-demo.jmx

7. Run the saved plan in non-GUI mode with a fresh result file.

   $ jmeter -n -t think-time-demo.jmx -l think-time-results.jtl -j jmeter.log
   Creating summariser <summary>
   Created the tree successfully using think-time-demo.jmx
   Starting standalone test @ 2026 Jun 30 21:25:13 GMT (1782854713448)
   Waiting for possible Shutdown/StopTestNow/HeapDump/ThreadDump message on port 4445
   summary =      4 in 00:00:06 =    0.7/s Avg:    73 Min:     0 Max:   290 Err:     0 (0.00%)
   Tidying up ...
   ... end of run

   -n runs JMeter without the GUI, -t selects the saved plan, -l writes sampler results, and -j writes the run log.
   Related: How to run a JMeter test from the command line

8. Check the result timestamps for the expected spacing.

   $ cat think-time-results.jtl
   timeStamp,elapsed,label,responseCode,responseMessage,threadName,dataType,success,failureMessage,bytes,sentBytes,grpThreads,allThreads,URL,Latency,IdleTime,Connect
   1782854714938,290,Checkout page request,200,OK,Checkout users 1-1,text,true,,16,0,1,1,null,0,0,0
   1782854716693,1,Checkout page request,200,OK,Checkout users 1-1,text,true,,16,0,1,1,null,0,0,0
   1782854717943,0,Checkout page request,200,OK,Checkout users 1-1,text,true,,16,0,1,1,null,0,0,0
   1782854719254,1,Checkout page request,200,OK,Checkout users 1-1,text,true,,16,0,1,1,null,0,0,0

   JMeter records sample timestamps as start times. The observed start gaps here are 1755 ms, 1250 ms, and 1311 ms; the gap includes the previous sampler's elapsed time plus the next timer delay.