The purpose of this study was to evaluate the influence of the medial longitudinal arch height on the shock wave that repetitively reaches the lower back in running. Impact forces were measured simultaneously at the ground by a force plate and at the level of the low back, by means of an accelerometer, skin-mounted at the L3 spinal process. The medial longitudinal arch height was calculated as navicular height divided by foot length. Twelve healthy subjects ran barefoot and with an identical sport shoe at a constant speed. The sample size was divided equally into a low-arch and a high-arch group. Statistical analysis was performed by multivariate analysis of variance and Pearson's correlation. At low back level, there was a significantly lower acceleration amplitude and rate in the high-arch group (amplitude = mean, 1.74 g and SD, 0.94 g; rate = mean, 71.2 g/sec and SD, 58.0 g/sec) compared with the low-arch group (amplitude = mean 2.25 g and SD, 1.11 g; rate = mean, 111.5 g/sec and SD, 68.6 g/sec) (P < 0.001, each). At the ground, there was a slight negative correlation between arch height and initial loading rate in AP (-0.19; P < 0.01) and vertical (-0.22; P < 0.001) directions and a positive correlation between arch height and initial loading rate in the medial direction (0.22, P < 0.05). The results indicate that the high-arch foot is a better shock absorber with regard to the low back level than the low-arch foot.