This study aimed to explore the limitations of the Ashworth scale for measuring spasticity. An isokinetic dynamometer to quantify resistance to passive stretch and surface EMG was used to verify if a stretch response occurred and, if so, at what joint angle. The authors sought to determine which components of passive resistance (magnitude, rate of change, onset angle of stretch, or velocity dependence) were most related to Ashworth scores and which were related to motor function in cerebral palsy (CP). Twenty-two individuals with spastic CP (11 males, 11 females; mean age 11.9 years, SD 4.3) and a comparison group of nine children without CP (four males, five females; mean age 11.3 years, SD 2.5) participated in the study. The group with CP included those with a diagnosis of spastic diplegia, hemiplegia, or quadriplegia, distributed across Gross Motor Functional Classification Levels. Procedures included: (1) clinical assessment at the knee joint, (2) functional assessments, and (3) isokinetic assessment of passive resistance torque in hamstrings and quadriceps at three velocities. EMG data were recorded simultaneously to identify stretch responses. Detecting stretch responses using the Ashworth scale compared with instrumented measures showed near complete agreement at extremes of the scale, with marked inconsistencies in mid-range values. Ashworth scores were correlated with instrumented measures, particularly for the quadriceps, with higher correlations to the rate of change in resistance (stiffness) and onset angle of stretch than to peak resistance torque. Those with greater resistance tended to have poorer function with isokinetic relations typically stronger.