Endurance runners are often advised to use 90 strides min(-1), but how optimal is this stride frequency and why? Endurance runners are also often advised to maintain short strides and avoid landing with the feet too far in front of their hips or knees (colloquially termed 'overstriding'), but how do different kinematic strategies for varying stride length at the same stride frequency affect economy and impact peaks? Linear mixed models were used to analyze repeated measures of stride frequency, the anteroposterior position of the foot at landing, V̇O2 , lower extremity kinematics and vertical ground reaction forces in 14 runners who varied substantially in height and body mass and who were asked to run at 75, 80, 85, 90 and 95 strides min(-1) at 3.0 m s(-1). For every increase of 5 strides min(-1), maximum hip flexor moments in the sagittal plane increased by 5.8% (P<0.0001), and the position of the foot at landing relative to the hip decreased by 5.9% (P=0.003). Higher magnitudes of posteriorly directed braking forces were associated with increases in foot landing position relative to the hip (P=0.0005) but not the knee (P=0.54); increases in foot landing position relative to the knee were associated with higher magnitudes (P<0.0001) and rates of loading (P=0.07) of the vertical ground reaction force impact peak. Finally, the mean metabolically optimal stride frequency was 84.8±3.6 strides min(-1), with 50.4% of the variance explained by the trade-off between minimizing braking forces versus maximum hip flexor moments during swing. The results suggest that runners may benefit from a stride frequency of approximately 85 strides min(-1) and by landing at the end of swing phase with a relatively vertical tibia.