The characteristics of the midsole were examined in four pairs of running shoes by a materials test. The variables of interest were the peak acceleration, time to peak acceleration and the kinetic energy absorbed. Ten subjects then ran at a recreational jogging pace (3.5 ms-1) barefoot and in the shoes. An accelerometer secured to the lower tibia was used to measure the peak acceleration and time to peak acceleration associated with footstrike. Subjects were also videoed and a kinematic analysis was undertaken at the knee and ankle joints. The results from the materials test showed that the shoes differed in their midsole characteristics, however, no significant differences (P > 0.05) were observed in the peak acceleration and time to peak acceleration during running in shoes. These variables were significantly greater in the barefoot running condition (P < 0.05), as compared with running in shoes. Small and subtle kinematic differences were observed between the barefoot and shoe conditions. It appears that the differences observed between the shoes in the materials test were not sufficient to elicit the kinematic changes observed between the barefoot and shoe conditions. It is suggested that runners operate within a 'kinetic bandwidth' when responding to impact stresses.