We compared a single perioperative cycle of adjuvant combination chemotherapy with no adjuvant treatment in a randomized trial (Ludwig Trial V) including 1275 patients with breast cancer who had no axillary-node metastases. The chemotherapy was administered on days 1 and 8, beginning within 36 hours after mastectomy, and consisted of cyclophosphamide, methotrexate, fluorouracil, and leucovorin. At a median follow-up of 42 months, the mean four-year disease-free survival (+/- SE) was 77 +/- 2 percent among the patients who received chemotherapy perioperatively, as compared with 73 +/- 2 percent among the patients who received no adjuvant treatment (hazard ratio, 0.77; 95 percent confidence interval, 0.61 to 0.98; P = 0.04). An advantage was observed for both premenopausal and postmenopausal women. The magnitude of the treatment effect was largest among patients with no or low estrogen-receptor content in the primary tumor. We conclude that chemotherapy modifies the post-operative course of node-negative breast cancer. Further trials to investigate an optimal selection of patients and treatments should be regarded as the best available therapeutic approach.