The aim of the present study was to investigate whether current smoking, ex-smoking, or amount of current smoking among postmenopausal women was associated with plasma levels of testosterone, and rostenedione, and DHEAs. About 65,000 women in Sweden have participated in two population-based prospective cohort studies where blood samples were collected at baseline, and information on life-style, reproductive history and anthropometrical measurements were available. The present study was based on 407 control subjects from a previous nested case-control study on the relation between steroid hormone levels and risk of breast cancer. A multivariate logistic regression analysis, adjusted for potential confounders, was used in order to obtain odds ratios (OR) with 95% confidence intervals. There was a high risk of high testosterone levels (above the median) in current vs. never smokers, OR: 1.85 (1.06-3.23). Risk of high testosterone levels increased by amount of daily smoking (increments of 10 cigarettes/day), OR: 1.55 (1.02-2.37). Ex-smoking was associated with high testosterone levels, OR: 1.56 (0.98-2.47). Current smoking and increasing amount of current smoking were weakly associated with high androstenedione levels. However, these associations did not reach statistical significance. No association was seen between smoking habits and DHEAs levels. We conclude that current smoking, and increasing amount of daily smoking, is associated with high testosterone levels.