A cyclical, sequential combination of tamoxifen and megestrol acetate (group B) was compared with conventional therapy (tamoxifen alone, group A) in 261 breast cancer patients. There was no statistically significant difference between groups for overall response rate (complete+partial response: group A, 35.8%, group B, 40.8%; p = 0.505) or for median response duration in responders (group A, 128 weeks, group B, 136 weeks; p = 0.488). Median survival from randomization was longer in those patients receiving sequential therapy (group A, 90 weeks, group B, 134 weeks) with a significantly lower relative death rate (group B/group A = 0.67; p = 0.011). This survival benefit appears to be due to a delay in progression among nonresponders in the sequential therapy group.