Objective: To assess the relationship between objective measures of endurance capacity and the self-report measures of physical function derived from the SF-36 Health Status Questionnaire. Design: Descriptive comparison study. Setting: Postacute residential treatment center. Participants: The sample consisted of 32 persons recovering from traumatic brain injury (TBI) admitted to a postacute residential rehabilitation facility. A total of 29 males and 3 females gave written consent to participate. Subjects were excluded if they were unable to follow 2-step commands and ambulate independently without manual assistance. Intervention: Not applicable. Main outcome measures: Subjects either walked on a treadmill until exhaustion during which peak aerobic capacity (VO₂) was measured (n = 32) or walked overground for 6 minutes (6 MW) during which intensity and total distance walked were monitored (n = 24). In addition, each subject completed the SF-36 Health Status Questionnaire and correlations between measures were calculated. Results: The relationship between total distance on the 6 MW and SF-36 measures of physical function (PF), vitality (VT), and the physical component summary (PCS) score were 0.44, 0.46, and 0.41 (P < .05), respectively. The relationship between peak VO₂ and PF, VT, and PCS was 0.17, 0.16, and 0.15, respectively. Conclusions: Self-reports of physical function obtained by the SF-36 may not be valid estimates of endurance capacity in patients with TBI during postacute rehabilitation. The SF-36 questions related to physical function are more general than the measure of endurance obtained by incremental treadmill walking (peak VO₂) and may account for the poor relationship. Overground walking (6 MW total distance) was better correlated to self-reports of physical function in patients with TBI.