The estrogenic effect of butylparaben was investigated in a rainbow trout Oncorhynchus mykiss test system. Butylparaben was administered orally to sexually immature rainbow trout every second day for up to 10 days in doses between 4 and 74 mg kg(-1) 2 d(-1) and in the water at 35 and 201 microgram l(-1) for 12 days. Plasma vitellogenin was measured before and during the exposures and the concentrations of butylparaben in liver and muscle were determined at the end of experiments. Increases in average plasma vitellogenin levels were seen at oral exposure to 9 mg butylparaben kg(-1) 2 d(-1). The ED50 values for increase in vitellogenin synthesis were 46, 29 and 10.5 mg butylparaben kg(-1) 2 d(-1), respectively, at day 3, 6 and 12. Exposure to 201 microgram butylparaben l(-1) increased vitellogenin synthesis, but exposure to 35 microgram l(-1) did not. Butylparaben showed little tendency to bioaccumulation in rainbow trout; less than 1 per thousand of the total amount of butylparaben administered orally at 51 mg kg(-1) 2 d(-1) over the 12 days experimental period was retained in liver at the end of the experiment. After 12 days exposure to 35 and 201 microgram butylparaben l(-1), plasma concentrations were 9 and 183 microgram l(-1), respectively, and for the fish exposed to 201 microgram l(-1) there was a positive correlation between concentrations of vitellogenin and butylparaben in the plasma. On the assumption that butylparaben removed from the water phase during water exposure were taken up into the fish, butylparaben uptake rates in the fish exposed to 35 and 201 microgram butylparaben l(-1) were 13 and 78 mg kg(-1) day(-1), respectively.