Spinal cord injury (SCI) causes inactivation and consequent unloading of affected skeletal muscle and bone. This cross-sectional study investigated correlations of muscle and bone in spinal cord-injured subjects compared with able-bodied subjects. Thirty-one complete SCI paraplegics were divided according to the neurological level of injury (NLoI) into group A (n=16, above thoracic 7 NLoI, age: 33±16 years, duration of paralysis (DoP): 6±6 years) and group B (n=15, thoracic 8-12, age: 39±14 years, DoP: 5.6±6 years), compared with 33 controls (group C). All were examined with peripheral quantitative computed tomography at 66% of tibia length (bone and muscle area, bone/muscle area ratio). In able-bodied subjects, muscle area was correlated with bone area (P<0.001, r=0.88). Groups A and B differed significantly from the control group in terms of bone and muscle area (P<0.001). In paraplegics, less muscle per unit of bone area (bone/muscle area ratio) was found compared with controls (P<0.001). Bone area was negatively correlated with the DoP in the total paraplegic group (r=-0.66, P<0.001) and groups A and B (r=-0.77, P=0.001 vs r=-0.52, P=0.12, respectively). Muscle area and bone/muscle ratio area correlations in paraplegic groups with DoP were weak. Paraplegic subjects who performed standing and therapeutic walking had significantly higher bone area (P=0.02 and P=0.013, respectively). The relationship between bone and muscle was consistent in able-bodied subjects and it was predictably altered in those with SCI, a clinical disease affecting bone and muscle.