Objective: To explore the relationships between weight-bearing, pain and walking velocity in the first four weeks of prosthetic training. Subjects: Consecutively referred transtibial amputees (n = 27). Design: Longitudinal study. Setting: Inpatient rehabilitation centre. Interventions: Measurements of static weight-bearing (SWB), perceived pain during the SWB test and self-selected walking velocity (VEL) were taken at entry, and at four weeks into the prosthetic training programme. Path analysis was conducted using a series of linear regression analyses to determine predictors of walking velocity at week 4. Results: Significant improvements from week 1 to week 4 were found in SWB, reported pain and velocity. SWB at week 4, when coupled with age, predicted 66% of the variance in walking velocity at week 4. Pain did not directly predict velocity. However, SWB at entry and perceived pain at week 4 predicted 56% of the variance in SWB at week 4. Conclusions: These analyses suggested a model in which SWB and age were both significantly related to velocity, and further, that pain played a role in velocity through a relationship mediated by its direct effect on SWB. Unlike age, SWB is a predictor of velocity that might be altered through pain reduction, wound-healing and rehabilitative interventions.