The purpose of this study was to determine the factors associated with flap-specific complications in patients receiving gracilis myocutaneous flaps for pelvic reconstruction, specifically to determine whether the short gracilis flap increases the risk of major flap loss. Records of consecutive women undergoing gracilis flap pelvic reconstruction procedures were analyzed retrospectively. Twenty-two received classical "long" flaps, while 24 received short flaps. In the short gracilis flap modification, the dominant vascular pedicle is deliberately sacrificed. Clinical factors for association with flap-specific complications were analyzed. Eighty-three flaps were used in 46 women for reconstruction after radical pelvic surgery. Those receiving short flaps were more likely to have a preoperative diagnosis of cervical cancer and to have undergone exenteration with rectosigmoid anastomosis, and all short flaps were performed by gynecologic oncologists (all P < 0.05). Distribution of other patient characteristics were similar. Loss of more than one or both flaps occurred in seven (15%) patients, three (14%) with long and four (17%) with short flaps, a nonsignificant difference. Age, weight, tobacco use, associated medical conditions, diagnosis, prior therapy, type of procedure, previous lymphadenectomy, use of bilateral flaps, postoperative abscess/fistula, and type of surgeon were not significantly associated with major flap loss. Four (50%) of eight patients undergoing exenteration with rectosigmoid anastomosis had major flap loss, compared to 2 of 29 with exenteration only, and 1 of 9 with other pelvic reconstructions (P < 0.05). In conclusion, sacrifice of the dominant vascular pedicle in the short gracilis flap does not increase the risk of major flap loss. Vaginal reconstruction after exenteration with rectosigmoid anastomosis significantly increases the risk of this complication.