Although a number of studies compare different techniques of breast reconstruction there seems to be a paucity of information documenting the factors that affect breast shape and symmetry after immediate reconstruction. A photographic analysis by 5 plastic surgeons (who were blinded to the nature of the procedure) of 62 patients undergoing skin-sparing mastectomy and immediate reconstruction was undertaken in an endeavor to identify these factors. Autologous techniques used included deepithelialized pedicled transverse rectus abdominis musculocutaneous (TRAM) flaps (n = 23; in 9 patients a bilateral breast reduction using inverted-T keyhole skin markings was performed, and the TRAM flap was used to fill the reduced skin pocket) and 4 patients had extended latissimus dorsi flaps. Prosthetic reconstruction was undertaken in 35 patients in whom a bilateral breast reduction skin pattern was used in 23 (prosthesis inserted submuscularly in 5 patients and subcutaneously in 18 patients) or, in another 12 patients the skin envelope was retained and the prosthesis was inserted partially submuscularly in 7 patients and subcutaneously in 5 patients. As a total group, the results of autologous reconstruction were better than prosthetic reconstruction (P = 0.048, Mann-Whitney U test). However, when a breast reduction pattern was used and the prosthesis was inserted subcutaneously, the results were not significantly better than when the reduced skin envelope was filled by autologous tissue (P = 0.64, Mann-Whitney U test). Failure to replace the areola leads to a smaller reconstructed breast and asymmetry. The skin envelope, when retained in toto, seems to be a major factor affecting breast shape. In patients with a reduction pattern applied, the skin envelope again seems to be a major factor affecting shape, because the results are similar whether the envelope was filled with autologous tissue or prosthetic material. The neoparenchyma is an important factor affecting breast projection. In patients undergoing prosthetic reconstruction, the results were significantly better with subglandular placement compared with submuscular placement (P = 0.007, Mann-Whitney U test). Patients who developed a complication (requiring debridement in the operating room or that took more than a month to achieve healing) had a poorer cosmetic result (P = 0.015, Mann-Whitney U test). These factors need to be considered when planning breast reconstruction to optimize the aesthetic result.