Twenty nonsymptomatic subjects were assessed while walking at a photoelectronically monitored place (2 +/- 0.1 m.s-1) using high speed cinematography (200 Hz) to record the rearfoot motion in the frontal plane, and electrogoniometry (100 Hz) to measure joint kinematics in the lower extremity. The foot type of the subjects was determined statically by using a podiascope and digitization techniques. The results demonstrated that no foot type variables contributed significantly to the variance in either rearfoot angle at foot strike or maximum rearfoot angle (p > 0.05). Regression equations were developed using kinematic variables: rearfoot angle at foot strike = 3.81 + (0.06*time to hip internal rotation) - (0.46*tibia internal rotation) + (0.14*plantarflexion); (R = 0.87, SE = 1.23 degrees); maximum rearfoot angle = 4.02 + (0.52*hip internal rotation) - (0.11*time to hip internal rotation); (R = 0.66, SE = 2.07 degrees). This study identifies hip joint movements as being the most significant contributors to prediction of rearfoot angles produced during walking.