The controversy over whether breast parenchyma should be approximated by suture after excisional biopsy is largely unresolved. The major debate is whether suture closure will decrease the incidence of hematomas and perhaps adversely alter the contour of the breast. To focus on the first question, we analyzed 162 patients undergoing mastectomy for breast cancer at Henry Ford Hospital between July 1988 and August 1992. All patients had previous excisional biopsies of breast lumps for diagnostic purposes. The cases were divided into two major groups depending on the placement or lack thereof of deep sutures in breast parenchyma. Of the total number of patients, 105 had deep suture approximation of breast parenchyma, whereas 57 did not. The biopsies were performed by 17 surgeons, although 72 per cent were done by four surgeons. Most of the surgeons varied their technique sporadically, but the number of cases most handled individually were too small for statistical analysis. Macroscopic and microscopic evaluation of the mastectomy specimen by the pathologist was used for definitive diagnosis of hematoma. The majority of these hematomas were of no clinical consequence and were usually not noted by the clinician. None of the hematomas required surgical drainage. Hematomas were determined pathologically in 64/162 patients (39.5%). Hematomas developed in 33/105 patients (31.4%) with deep wound approximation of breast parenchyma by suture compared to the higher rate seen in 31/57 patients (54.4%) without deep wound approximation (P = 0.004). This study supports the use of deep suture approximation after excisional biopsy for breast cancer.