It has been hypothesized that hepatocellular carcinoma might be a long-term adverse effect of tamoxifen therapy. Data from nine population-based cancer registries in the United States were used to investigate time trends in the incidence of hepatocellular carcinoma in white women previously diagnosed with invasive breast cancer during 1974-1987 and followed until 1989. Of particular interest were the rates after 1977, the year tamoxifen was licensed by the FDA. Compared to rates in all white women, no increased risk of hepatocellular carcinoma was found in women most likely to have received tamoxifen--those 50 years of age or more at diagnosis of breast cancer and diagnosed after 1977. These results suggest that tamoxifen does not cause a large increase in the incidence of hepatocellular carcinoma within the first decade after use. However, smaller and/or later increases in the risk for hepatocellular carcinoma are possible and warrant continued monitoring of women treated with tamoxifen.