Changes in thigh muscle volume over 6 months were assessed using magnetic resonance imaging in 11 subjects aged 6 to 47 years with spinal muscular atrophy (4 type 2 and 7 type 3; 4 ambulatory and 3 nonambulatory). Muscle volume with normal and abnormal signal was measured using blinded, semiautomated analysis of reconstructed data. Volumes at baseline and 6 months were correlated with clinical function at each epoch. There was minimal increase in normal (0.3 ± 1.4 mL/cm) and total (0.1 ± 1.3 mL/cm) muscle. Muscle volume correlated closely with clinical function. Minimal interval change in muscle volume is consistent with the established clinical history of minimal disease progression over intervals shorter than 1 year. Relative constancy of muscle volume estimation and correlation with established functional measures suggest a role for segmental magnetic resonance imaging as a biomarker of treatment effect in future therapeutic trials.