Breast reconstruction with transverse rectus abdominis muscle (TRAM) flap raises two contradictory questions: the vascular safety of the flap and the late abdominal wall sequellae. In order to analyse these sequellae, 71 patients with TRAM flap breast reconstruction at the Institut Curie had a late postoperative evaluation by both a physiotherapist and a surgeon, an average 28 months after their reconstruction. 12 had had a double pedicled TRAM (DPT) and 59 a single pedicled TRAM (SPT). Hernias and bulges were systematically recorded, and all patients had an evaluation of their abdominal wall function by questioning (subjective evaluation) and muscular testing (objective evaluation). The overall hernia rate (including bulges) was 5.6%. This rate was 2.5% when mesh was used, and 9.5% when direct closure was performed. This hernia rate was not influenced by the type of TRAM (SPT or DPT). 20% of patients complained of residual abdominal pain, and 36% of a decrease of their abdominal strength after SPT. Both these figures were 75% after DPT. Testing showed that these sequellae were related to an impairment of the supraombilical portion of the rectus, this impairment being much higher after DPT than SPT: none of the 12 patients with DPT were able, from a lying position, to sit down without using their hands (not reaching 4 in Lacote's test), whereas 47% of the SPT could do it. The oblique muscles were also impaired, as less than 20% of patients reached Lacote 4. However, this impairment was not influenced by the type of flap harvested. Testing was also equivalent after both techniques of SPT (standart or "supercharged"). The post-operative hernia rate was not higher for DPT and seemed related to the technique used for abdominal wall closing (mesh vs direct closure). However, the functional sequellae (pain, muscle strength decrease) were much higher after DPT than SPT. It thus confirms us in our attitude to restrict the indications of DPT, when feasible, to the profit of microsurgical flaps.