A retrospective cohort-study in 4109 breast cancer patients was undertaken to determine how tamoxifen affected the risk of endometrial cancer. Data on 1701 tamoxifen-treated women were analysed. Two thousand four hundred and eight non-tamoxifen users served as control group. The occurrence of new primary uterine cancers was assessed by computerized linkage to the Austrian Cancer Registry. Twenty-five women who subsequently developed endometrial cancer were identified. Eight uterine cancers occurred in the tamoxifen group, whereas 17 uterine cancers were found in the control group. The estimate of the relative risk (RR) showed an increased risk to develop endometrial cancer for the tamoxifen group RR 1.136 (95% CI 0.71; 1.80). Analysis of relevant confounding variables did not show any differences in the two groups. In conclusion, this retrospective study demonstrated a non-significant increased risk of endometrial cancer in women receiving tamoxifen as treatment for breast cancer. However, the magnitude of RR and the absolute number of endometrial cancer cases in this long term observation demonstrate clearly that the clinical benefit of tamoxifen therapy greatly outweighs the risk.