The pharmacokinetics of methanol and formate were characterized in male Fischer-344 rats and rhesus monkeys exposed to methanol vapor concentrations between 50 and 2000 ppm for 6 hr. End-of-exposure blood methanol concentrations were not directly proportional to the atmospheric concentration. The methanol exposures did not cause an elevation in blood formate concentrations. After an intravenous dose of [14C]methanol in rats, metabolism, exhalation, and renal excretion contributed 96.6, 2.6, and 0.8%, respectively, to the elimination of blood methanol concentrations. These values and the calculated renal methanol extraction efficiency (0.007) are nearly identical to those for humans after low doses of methanol. A physiologically based pharmacokinetic model was developed to simulate the in vivo data. In order to simulate the observed blood methanol concentrations in the inhalation studies in rats, a double pathway for methanol metabolism to formaldehyde was used. One path used rodent catalase Km and Vmax values and the other used a smaller Km and Vmax to simulate an enzyme with a higher affinity and lower capacity. The lack of proportionality observed in end-of-exposure blood methanol concentrations may be due to saturation of an enzyme with higher affinity and lower capacity than catalase. The physiologically based pharmacokinetic model was modified to simulate the monkey data and was scaled-up for humans. In order to simulate the monkey blood methanol concentrations, the use of rodent catalase parameters for methanol metabolism was required. This finding suggests that primates and rodents may be similar in the initial step of methanol metabolism after low methanol doses. Previously published human urinary methanol excretion data was successfully simulated by the model. The models were used to predict the atmospheric methanol concentration range over which the laboratory species exhibit quantitative similarities with humans. Below 1200 ppm, all three species exhibit similar end-of-exposure blood methanol concentrations and a linear relationship between atmospheric and blood methanol concentrations. At higher atmospheric concentrations, external and internal methanol concentrations increase desparately, suggesting that delivered dose rather than exposure concentration should be used in interpreting data from high-dose studies.