Between 1978 and 1981 we conducted a trial in which adjuvant endocrine therapy consisting of tamoxifen (T = 20 mg/d) and low-dose prednisone (p = 7.5 mg/d) for the duration of one year (p + T), was compared with no adjuvant therapy (observation) in 320 women with operable breast cancer aged 66 to 80 years (median age, 70 years). All patients had axillary lymph node metastases after at least a total mastectomy and axillary clearance. At 96 months median follow-up, 9.1% of the patients died without apparent relapse from cancer. An additional 1.9% had a second malignant neoplastic disease (not breast cancer). The 8-year disease-free survival (DFS) percentages (+/- SE) for the p + T and the observation groups were 36% (+/- 4%), and 22% (+/- 3%), (P = .004). The 8-year overall survival percentages were 49% (+/- 4%) and 42% (+/- 4%), respectively (P = .43). We conclude that despite a large proportion of deaths without relapse of breast cancer, a significant advantage for the p + T group in terms of DFS was demonstrated. We hypothesize that an endocrine therapy of longer duration might have an overall survival benefit in a population of elderly patients.