The purpose of this study was to determine the effects of systematic changes in stride rate and length at a given running speed on the peak shank deceleration (PSD) experienced during ground contact. Data were collected from 10 well-trained subjects as they ran on a treadmill at a pace of 3.8 m s-1 (7-min mile-1). Shank deceleration was measured by a lightweight accelerometer which was tightly attached over the distal medial tibia. High-speed films (200 Hz) were taken from a side view to quantify modifications in sagittal plane movement which might have accompanied the stride rate changes. Five stride rate conditions were randomly presented - 10% slower, 5% slower, normal, 5% faster and 10% faster. Average PSD was computed from 10 steps during each condition and testing was repeated on three different occasions. For each session, PSD observed for each condition was normalized to that observed at the normal stride rate in order to minimize the effects of variations in attachment of the accelerometer between and within subjects. The normalized PSD results at each stride rate tested were - normal = 1.0, 10% slower = 1.09, 5% slower = 1.03, 5% faster = 0.96 and 10% faster = 0.91. Significant differences were found between all these means except normal and 5% slower. The kinematic analysis revealed non-significant results for hip, knee and ankle joint angles at touchdown for the various stride rates. Application of the findings to existing analytical models indicated that, for a given running speed, peak impact forces in the ankle and knee joints decreased as stride rate increased.