Spinal cord injury (SCI) is characterized by marked bone loss at the knee, and there is a need for established dual-energy X-ray absorptiometry (DXA) protocols to examine bone mineral density (BMD) at this location to track therapeutic progress and to monitor fracture risk. The purpose of this study was to quantify the precision and reliability of a DXA protocol for BMD assessment at the distal femur and the proximal tibia in individuals with SCI. The protocol was subsequently used to investigate the relationship between BMD and duration of SCI. Nine individuals with complete SCI and 9 able-bodied controls underwent 3 repeat DXA scans in accordance with the short-term precision methodology recommended by the International Society of Clinical Densitometry. The DXA protocol demonstrated a high degree of precision with the root-mean-square standard deviation ranging from 0.004 to 0.052 g/cm and the root-mean-square coefficient of variation ranging from 0.6% to 4.4%, depending on the bone, the region of interest, and the rater. All measurements of intra- and inter-rater reliability were excellent with an intraclass correlation of ≥0.950. The relationship between the BMD and the duration of SCI was well described by a logarithmic trend (r = 0.68-0.92). Depending on the region of interest, the logarithmic trends would predict that, after 3 yr of SCI, BMD at the knee would be 43%-19% lower than that in the able-bodied reference group. We believe the DXA protocol has the level of precision and reliability required for short-term assessments of BMD at the distal femur and the proximal tibia in people with SCI. However, further work is required to determine the degree to which this protocol may be used to assess longitudinal changes in BMD after SCI to examine clinical interventions and to monitor fracture risk.