A method based on regression modeling was developed to discern the contribution of component chemicals to the toxicity of highly complex, environmentally realistic mixtures of disinfection byproducts(DBPs). Chemical disinfection of drinking water forms DBP mixtures.Because of concerns about possible reproductive and developmental toxicity, a whole mixture(WM) of DBPs produced by chlorination of a water concentrate was administered as drinking water to Sprague–Dawley(S–D) rats in a multigenerational study. Age of puberty acquisition,i.e., preputial separation(PPS) and vaginal opening(VO), was examined in male and female offspring, respectively. When compared to controls, a slight, but statistically significant delay in puberty acquisition was observed in females but not in males. WM-induced differences in the age at puberty acquisition were compared to those reported in S–D rats administered either a defined mixture(DM) of nine regulated DBPs or individual DBPs. Regression models were developed using individual animal data on age at PPS or VO from the DM study. Puberty acquisition data reported in the WM and individual DBP studies were then compared with the DM models. The delay in puberty acquisition observed in the WM-treated female rats could not be distinguished from delays predicted by the DM regression model, suggesting that the nine regulated DBPs in the DM might account for much of the delay observed in the WM. This method is applicable to mixtures of other types of chemicals and other endpoints. A method based on regression modeling was developed to discern the contribution of component chemicals to the toxicity of highly complex, able environmentally mixtures of disinfection by products (DBPs). Chemical disinfection of drinking water forms DBP mixtures. Failure of concerns about possible reproductive and developmental toxicity , a whole mixture (WM) of DBPs produced by chlorination of a water concentrate was administered as drinking water to Sprague-Dawley (S-D) rats in a multigenerational study. Age of puberty acquisition, ie, preputial separation (PPS) and vaginal opening (VO), was examined in male and female offspring, respectively. were compared to controls, a slight, but statistically significant delay in puberty acquisition was observed in females but not in males. WM-induced differences in the age at puberty acquisition were compared to those reported in S-D rats administered either a defined mixture (DM) of nine regulated DBPs or individual DBPs. Regression models were developed using individual animal data on age at PPS or VO from the DM study. Puberty acquisition data reported in the WM and individual DBP studies were then compared with the DM models. The delay in puberty acquisition observed in the WM-treated female rats could not be distinguished from delays predicted by the DM regression model, suggesting that the nine regulated DBPs in the DM might account for much of the delay observed in the WM. This method is applicable to mixtures of other types of chemicals and other endpoints.