Abdominal-wall function was evaluated preoperatively and at intervals postoperatively in 25 consecutive patients undergoing breast reconstruction with transverse rectus abdominis myocutaneous (TRAM) flaps (single-pedicled TRAM flap, 14 patients; free TRAM flap, 9 patients; and bilateral free TRAM flaps, 2 patients). Objective measures of abdominal-wall function were performed with the B200 Isostation, a triaxial dynamometer. In addition, the patients were assessed by a physical therapist and filled out an activity questionnaire at each postoperative examination. Tests of abdominal-wall function demonstrated the greatest decrease in performance at the 6-week postoperative tests of flexion. The maximum isometric flexion torque of the pedicled TRAM flap group decreased to 58 +/- 10 percent, while the unilateral free TRAM flap group average was 87 +/- 11 percent of baseline. For the pedicled TRAM flap group this difference was significant (p = 0.004). By the 6-month evaluation, the maximum isometric flexion torque increased for both the pedicled and the free TRAM flap groups to 89 +/- 13 percent and 93 +/- 8 percent of baseline, respectively. The physical therapist evaluation of abdominal-wall strength and the activity questionnaire data showed no statistically significant differences between groups or over time. Rectus abdominis muscle harvest for pedicled TRAM flaps causes a greater insult to the abdominal wall than does free TRAM flap harvest. The ultimate clinical effect of the sacrifice of even an entire rectus abdominis muscle appears to be well tolerated by most patients. This is the first prospective outcome study of abdominal-wall function in TRAM flap patients. The clinical implications of this information will be discussed.