The tonic stretch reflex threshold in children with cerebral palsy (CP) was measured to determine its test–retest reliability and its concurrent validity as a potential measure of spasticity. Fourteen children with spastic CP aged 6 to 18 years were tested on three separate occasions for clinical spasticity and stretch reflex thresholds of affected elbow flexors. Electromyographic (EMG) recordings were obtained by surface electrodes for elbow flexors and extensors during mechanical displacements of the passive joint towards extension. Displacements were produced by a torque motor at seven velocities which randomly varied from trial to trial. EMG activity was measured in the stretched flexor muscles to determine threshold angles and velocities for each velocity of stretch. These were plotted on a velocity–angle-phase diagram and regression analysis was used to determine the static stretch reflex threshold for each participant. The measure showed good test–retest reliability for the group (ICC 0.73, p<0.001) whereas a significant correlation between the measure and the clinical spasticity scale was not found. This technique is a potential outcome variable for measuring the efficacy of treatments aimed at decreasing spasticity in children with CP.