The disposition of antipyretic drugs is central to the understanding of their action in children. Accordingly, the authors measured plasma levels of acetaminophen and ibuprofen in 153 febrile children for 6 hours after a single dose of either acetaminophen (12.5 mg/kg) or ibuprofen (5 or 10 mg/kg). C(max) occurred about 2 1/2 hours before maximum antipyresis, when plasma acetaminophen or ibuprofen was 25 to 50% less than C(max). Most plasma level data fit a one-compartment open model, and this suggests a pharmacodynamic basis for the observed lag between C(max) and maximum antipyretic response. Plasma levels (and AUC(O --> infinity)) of ibuprofen 10 mg/kg were less than expected for a two-fold increase in dose. For acetaminophen, the t(lag) was less than ibuprofen, K(a) was more than ibuprofen, and beta was less than ibuprofen. The ibuprofen beta was not dose dependent, but the V(d) was dose and model dependent. In contrast, ibuprofen Cl(p) was dose and model independent. Acetaminophen pharmacokinetics were similar to those previously reported. Initial temperature, race, gender, prior medications, or diagnosis did not confound the results for ibuprofen or acetaminophen. Accordingly, a pharmacodynamic basis is a more likely explanation for the initial temperature effects found previously for antipyretic drugs in children. Ibuprofen (5 and 10 mg/kg) AUC(O --> infinity) was higher in the older (greater-than-or-equal-to 2.5 yrs) children and the V(d) and Cl(p) were lower in the older children, when discriminated by age or pharmacokinetic parameters. The observed dose dependency of AUC(O --> infinity) and the effect of age on ibuprofen disposition must be considered if pharmacokinetic interpretations are used to develop the antipyretic dose of ibuprofen in children.