To date, no five year survival rates have been reported for patients with hepatic metastases (HM) from breast origin treated by chemotherapy or hormone therapy. This study was done to evaluate whether or not surgical excision of such metastatic disease associated with chemotherapy has any effect on such a poor prognosis. Between May 1985 and September 1988, 22 patients with a diagnosis of isolated (solitary or multiple) HM have been surgically treated. The therapeutic approach also included systemic preoperative and postoperative chemotherapy. Laparotomy findings for those 22 patients are presented. There were benign hepatic lesions in four patients (four different histologic types) and hepatectomy was done in two patients. Diffuse metastatic disease contraindicated any attempt at hepatic resection in six patients. Hepatectomy was possible in 12 patients. There was no postoperative mortality and minimal morbidity. One-half of the patients with metastases had one or more positive lymph nodes at the hepatic pedicule. Median survival time for 12 patients treated by hepatectomy was 37 +/- 9 months after hepatectomy and 42 +/- 3 months after the discovery of HM. Eleven patients had other metastases after an average period of 11.8 months after hepatectomy, and the liver was the first site of recurrence in eight. One patient underwent a hepatectomy twice. Two patients were free of disease 29 to 46 months postoperatively in spite of poor prognostic features on the excised specimen. Although the median survival time was two times better than with usual treatment, we believe this procedure is of doubtful benefit. In this situation, hepatectomy acts as cytoreductive surgical treatment, with the main limiting factor being the relatively low effectiveness of chemotherapy. This type of therapy can be applied only to a restricted number of patients, and if we decide to proceed with this study, we will have to modify the protocol of chemotherapy radically.