The purpose of this study was to examine the influence of short hamstring muscles (hamstrings) on gravitational torque of the leg (GT) and on knee extension and knee flexion concentric isokinetic torque. Fifteen non-disabled men (ages: 18 to 37 years) with clinically short hamstrings(passive straight-leg-raising ≤ 65°) were measured for the GT in two positions: 1) seated position that included a passive stretch of the hamstrings, and 2) supine position that did not include a passive stretch of the hamstrings. Subjects performed three trials of maximal concentric knee extension and flexion in the seated position under four randomly ordered conditions(60°·s
$^{-1}$
with seated GT, 60°·s
$^{-1}$
with supine GT, 180°·s
$^{-1}$
with seated GT, and 180°·s
$^{-1}$
with supine GT). One-way ANOVA analysis showed that the GT measured while seated(36.00 ± 9.58 Nm) was greater than the GT measured while supine (21.00 ± 3.90 Nm) (P < 0.001). The greater GT while seated was explained by the sum of the passive hamstring torque and the true GT. Knee flexion peak torque, knee flexion average torque, and the hamstring/quadriceps average torque ratios were decreased at 60°·s
$^{-1}$
(P ≤ 0.023) using the seated GT compared to using the supine GT. The same torque characteristics did not differ significantly for knee flexion at 180°·s
$^{-1}$
, or for knee extension at either isokinetic speed. We concluded that hamstring passive torque contributed to greater GT in the seated position, which caused invalid knee flexion torque measurements when tested at 60°·s
$^{-1}$
. This evidence supports weighing the leg supine in order to remove hamstring passive torque and to ensure valid knee concentric isokinetic torque measurements when testing at 60°·s
$^{-1}$
.
