Kinematic and kinetic data were collected from 12 healthy subjects whi
lst they performed both downhill and level walking at a controlled cad
ence. A ramp of 6 m length and a gradient of -19% was used for downhil
l walking and this incorporated the same force platform that was used
for level walking. Planar net joint moments and mechanical power at th
e ankle, knee, and hip joints were calculated for the sagittal view us
ing force platform and video records based on standard inverse dynamic
s procedures. On the basis of differences in ankle, knee, and hip join
t kinematics the ankle joint was seen to compensate for the gradient a
t push off and during the swing, the knee joint from early stance thro
ugh until early swing phase, and the hip joint from early swing throug
h until the early stance phase. The major differences in joint moments
and muscle mechanical power were seen in the knee and ankle joint. Wh
ereas peak moments and muscle power were much higher for downhill walk
ing in the knee joint, these measures were significantly smaller at th
e ankle joint. Hip joint moments and muscle power estimates were only
slightly larger for downhill walking. These data explain well the prob
lems that patients with patellofemoral pathology and anterior cruciate
ligament (ACL) deficiency encounter with downhill walking, and the mu
scle soreness experienced by mountain trekkers. Relevance Biomechanica
l estimates of musculoskeletal loadings in gait are invariably derived
from laboratory studies of walking over a level surface. In this stud
y comparisons were made between downhill and level walking in order to
appreciate more fully the increased loadings on the lower extremity u
nder more stressful but not atypical conditions. The data so derived p
rovide the necessary basis for the prediction of loadings on specific
muscle/joint structures and can serve as a foundation for exercise pre
scription with patients recovering from injury or orthopaedic surgery.