Survival was studied in a population-based cohort of over 23,000 women who were prescribed hormone replacement therapy. Complete follow-up through 1986 revealed a total of 1472 deaths, which was somewhat lower than expected; the relative survival being 101.1% (95% CL, 100.8, 101.3) after 5 years and 102.4% (95% CL, 101.9, 102.8) after 10 years. The relative survival increased with increasing age at entry into the cohort, being 98.2% (95% CL, 96.6, 99.8) in the 40-44 and 105.2% (95% CL, 101.4, 109.1) in the 65-69 year age group after 10 years. Neither the type of compound (potent vs non-potent estrogens), nor the year of entry into the cohort seemed to affect survival, whereas survival advantage generally increased with years of follow-up. Multivariate analysis showed that age at time of first prescription was the only determinant that significantly affected the death risk. This pattern could be explained by confounding due to selection of healthy subjects receiving hormone replacement therapy and/or by the specific choice of estrogen compounds (and progestogens), related to age. It is concluded that hormone replacement therapy is associated with a survival which is similar to or--notably at ages above 50-60 years--slightly higher than that in the general population.