Purpose This prospective study examined the psychometric properties of the adapted Dutch translation of the Short Musculoskeletal Function Assessment (SMFA) questionnaire in patients with isolated unilateral lower fracture (LEF) or upper extremity fracture (UEF). Patients (N = 458) completed the SMFA, WHOQOL-BREF, and the RAND-36 at time of diagnosis (i.e. pre-injury status), 1, and 2 weeks post-fracture. Principal axis factoring was performed, and Cronbach's alpha coefficients (alpha) and intra-class correlation coefficients (ICC) were calculated. Furthermore, Pearson's product-moment correlations (r), paired t tests, and standardized response means (SRM) were calculated. A three-factor structure was found: Lower extremity dysfunction, Upper extremity dysfunction, and Daily life consequences. This structure was different for patients with LEF versus UEF. ICCs ranged from .68 to .90, and alpha varied from .81 to .95. The correlations between the SMFA and, respectively, the RAND-36 and WHOQOL-BREF were small to large depending on the SMFA factor combined with fracture location. Responsiveness was confirmed (p < .0001; SRM ranging from .28 to 1.71). The SMFA has good psychometric properties in patients with fractures. Patients with UEF and LEF could not be regarded as a homogenous group. The development of separate SMFA modules should be considered.