We designed a simple procedure based on the angular speed of the knee joint for quantitating the patellar tendon reflex. The angular speed of the knee joint is calculated from acceleration data generated in response to the tapping force applied to the patellar tendon with a customized tendon hammer and measured using a tri-axial accelerometer placed at the ankle joint. Data were collected and processed using a signal analyzer and a notebook PC. The results obtained using standard equipment were similar to those generated by more elaborate devices. For instance, the time delay (29.6+/-6.0 ms) and the acceleration time (150.8+/-19.5 ms) of the speed response were quite constant for all participants within the range of tapping forces normally applied during physical examinations. Representative relationships between the peak tapping force and the peak angular speed also closely fit with the exponential model (the average coefficient of determination, 0.70; range, 0.43-0.97). In contrast, the mean asymptotic value of the peak angular speed (Omega(pas)) was 160+/-67 degrees/s for spastic individuals, compared with only 72+/-21 degrees/s for healthy individuals. The important features of this method are portability, ease of use, and non-constraint of solicited reflex responses.