Substances identified as animal carcinogens are no longer used as ingredients of hair dyes. However, hair dyes are diverse groups of chemicals, and certain compounds may affect endogenous sex hormone levels. We examined the association between hair dye use and sex hormone levels among premenopausal women. Study subjects were 431 premenopausal Japanese women who had regular menstrual cycles less than 40 days long. Information on the use of hair dyes or hair bleach, the type of hair coloring used, the duration of use and the frequency of application was collected using a self-administered questionnaire. Fasting plasma samples were obtained to measure estradiol, testosterone, dehydroepiandrosterone sulfate, sex hormone-binding globulin, follicle-stimulating hormone and luteinizing hormone. After controlling for covariates, the mean plasma total testosterone level was about 14% higher in women who had used hair dyes for 10 or more years than that among women who had never used them (P for trend = 0.02). A similar association was observed when the type of hair dye was restricted to permanent hair dyes. A higher frequency of applying non-permanent hair dyes was marginally significantly associated with higher total and free estradiol levels. Data suggest that long-term use of hair dyes may be associated with an increase in circulating testosterone levels. As this is, to our knowledge, the first study examining the association between hair dye use and sex hormone levels, replication of the results is required.