The purpose of this investigation was to examine impacts during running. The protocol used to fulfill this purpose consisted of subjects running at a constant velocity but altering their stride frequency. In the first of three studies, shock attenuation was investigated as a possible optimizing criteria. It was found that impacts and shock attenuation both increased as the stride frequency decreased. The result was that shock at the head changed very little across stride frequency conditions. Because oxygen consumption and shock attenuation both increased during the low stride frequency conditions, it was thought that increasing the attenuation of shock in the body may have an associated energy cost. The second study examined the locations of muscular energy absorption during the impact phase of the running cycle. A rigid body model was used to estimate energy absorbed during the impact phase of running. During the low stride frequency conditions, the magnitude of the velocities of the support leg and the rest of the body were greater during ground contact. This required greater energy absorption at the hip, knee and ankle joints. Muscles that cross the knee joint adjusted the most in response to increased shock. It was conjectured that the perpendicular distance from the line of action of the resultant ground reaction force to the knee joint center played a role in this increased energy absorption. The third study used a mass-spring-damper model to simulate the vertical ground reaction forces of a human runner as stride frequency was altered. The input parameters of the model were measured from the actual characteristics of the runners when possible. Spring stiffness values were selected by an optimizing routine. An upper mass-spring system was used to control the active portion of the ground reaction force. The stiffness of this spring showed a twofold increase as the subjects increased their stride frequency. A lower, mass-spring-damper system was used to control the impact portion of the ground reaction force. This stiffness value showed a trend opposite that of the upper spring. The lower spring apparently prevented the support leg from completely collapsing during the low frequency running.