A mathematical model for the exchange of organic chemicals by fish gills was formulated based solely on the limitations imposed by the flows of water and blood into the gills. For large rainbow trout, the model was found to closely follow the magnitude and trends of observed gill uptake rates over a range of octanol/water partition coefficient from 1 to ten to the sixth power. Observations averaged only about 30% less than model predictions. The modest lack-of-fit is presumably due to the effects of diffusional barriers and ionization which would further limit uptake, although uncertainties in model parameters and data are likely also partly responsible. The analysis suggests that these basic physiologicalparameters are of major importance in the regulation of exchange at fish gills and should be accounted for in more detailed toxicokinetic models. Also, the simple model could by itself be useful for approximate assessments of accumulation of organic chemicals by fish.