Recent epidemiological studies on hormone replacement therapy (HRT) containing mainly conjugated equine estrogens indicate increased risk for venous thromboembolism (VTE). The purpose of the present epidemiological study was to evaluate the effect of HRT containing natural estrogens, i.e., estradiol, on the risk of VTE. HRT formulations containing estradiol are commonly used in Scandinavia. The study was a population-based case-control study. Cases were consecutive females, aged 44-70 years, discharged from Ullevål University Hospital with the diagnosis of deep venous thrombosis or pulmonary embolism during 1990-1996. Fifty-one women with cancer-associated thrombosis were excluded from the study. Controls were randomly collected from the same source population and matched by age. The material comprised 176 cases and 352 controls, i.e., 2 controls for each case. Only formulations containing estradiol were used. The frequency of HRT use was 28% (50/176) in cases and 26% (93/352) in controls. The estimated matched crude odds ratio with 95% confidence interval was 1.13 (0.71-1.78), which indicates no significant association of overall use of HRT and VTE. The estimated adjusted odds ratio was 1.22 (0.76-1.94) performing multi-confounder adjustment using the conditional logistic model and adjusting for hypertension, coronary heart disease, diabetes mellitus, smoking habit, BMI, and the presence of previous VTE. Stratification for time of exposure indicated an increased risk of VTE during the first year of use with a crude odds ratio of 3.54 (1.54-8.2). This effect was reduced by extended use to a crude odds ratio of 0.66 (0.39-1.10) after the first year of use. We conclude that use of HRT containing estradiol was associated with a threefold increased risk of VTE, but this increased risk was restricted to the first year of use.