Bone loss from the paralysed limbs after spinal cord injury (SCI) is well documented. Under physiological conditions, bones are adapted to forces which mainly emerge from muscle pull. After spinal cord injury (SCI), muscles can no longer contract voluntarily and are merely activated during spasms. Based on the Ashworth scale, previous research has suggested that these spasms may mitigate bone losses. We therefore wished to assess muscle forces after SCI with a more direct measure and compare it to measures of bone strength. We hypothesized that the bones in SCI patients would be in relation to the loss of muscle forces. Six male patients with SCI 6.4 (SD 4.3) years earlier and 6 age-matched, able-bodied control subjects were investigated. Bone scans from the right knee were obtained by pQCT. The knee extensor muscles were electrically stimulated via the femoral nerve, isometric knee extension torque was measured and patellar tendon force was estimated. Tendon force upon electrical stimulation in the SCI group was 75% lower than in the control subjects (p<0.01). Volumetric bone mineral density of the patella and of the proximal tibia epiphysis were 50% lower in the SCI group than in the control subjects (p<0.01). Cortical area was lower by 43% in the SCI patients at the proximal tibia metaphysis, and by 33% at the distal femur metaphysis. No group differences were found in volumetric cortical density. Close curvilinear relationships were found between stress and volumetric density for the tibia epiphysis (r(2)=0.90) and for the patella (r(2)=0.91). A weaker correlation with the tendon force was found for the cortical area of the proximal tibia metaphysis (r(2)=0.63), and none for the distal femur metaphysis. These data suggest that, under steady state conditions after SCI, epiphyseal bones are well adapted to the muscular forces. For the metaphysis of the long bones, such an adaptation appears to be less evident. The reason for this remains unclear.