A case-control study of the relationship between hair dye use and breast cancer included 129 breast cancer patients and 193 control women drawn from the records of a multiphasic screening clinic. Information was obtained by telephone interview on a number of risk factors for breast cancer and on variables pertaining to hair dye use: chronologic time, duration, frequency, type, and color. From this, quantitative measures of cumulative hair dye use at various intervals prior to breast cancer (or an equivalent for controls) were obtained. A multivariate risk factor score was used to control for confounding variables. The adjusted relative risks for breast cancer versus hair dye use were greater than unity but were not generally significant. However, integral measures of hair dye use (No. of yr times frequency per yr) were significantly related to breast cancer when confounding variables were controlled. The association between hair dye use and breast cancer was greatest among women over 50 years of age and among those at lower natural risk for breast cancer. An analysis of temporal patterns showed that breast cancer was related mainly to hair dye use 10 or more years before cancer diagnosis. Because of the retrospective nature of the hair dye data and the small sample size, these results require further validation.