A descriptive profile for antipyretic drug action has been documented for children. However, a linked pharmacokinetic-pharmacodynamic (PK/PD) model is central to the understanding of antipyretic drug action in febrile children. This was examined for previously reported data from 178 febrile children who received a single oral dose of acetaminophen (APAP) (12.5 mg/kg), ibuprofen (IBU) (5 or 10 mg/kg), or placebo. Rectal temperatures and plasma levels (microgram/ml) of APAP and IBU were measured for up to 12 hr after drug administration. Nonlinear regression analyses were applied to these measurements and yielded simultaneous solutions of an integrated one-compartment PK, link, and Sigmoid Emax effect model in 102/153 febrile children given APAP or IBU. The PK parameters (tlax, k alpha, beta, T1/2 beta, AUC0-infinity, Vd/F, and Clp/F) were not different than those reported previously, except the APAP k alpha was significantly lower. The link component yielded keos of 0.58 +/- 0.06 (mean +/- SE), 0.70 +/- 0.11 and 0.57 +/- 0.11 hr-1 for APAP, IBU05, and IBU10, respectively; the Sigmoid Emax component yielded EC50s (microgram/ml) and sigmoidicity (gamma) of 4.63 +/- 0.39 and 3.98 +/- 0.42 for APAP, 11.33 +/- 1.35 and 3.97 +/- 0.58 for IBU05 and 12.83 +/- 1.89 and 4.27 +/- 0.63 for IBU10. On visual inspection of the efficacy-time profiles of the febrile children, a number of them had an apparent linear function (slope; delta degree C/hr) and/or a sinusoidal cyclic function "confounding" standard approaches to PD analysis. Thus, the temperature profiles of 91/102 children given APAP or IBU required the addition of a slope (delta degree C/hr) and/or a sinusoidal cyclic function to the Sigmoid Emax component to fit the data satisfactorily. All 22 children given a placebo also required a slope and/or a cyclic function in their PD model. The residual delta degree Cs (observed-predicted) of the placebo group were not significantly different from 0. Thus, no placebo antipyretic effect was observed. Dose dependency of IBU AUC0-infinity was confirmed; doubling the dose from 5 to 10 mg/kg increased the AUC0-->infinity by only 1.5-fold. The confounding effect of initial temperature (Tempi) on antipyretic efficacy in all treatment groups except placebo was also confirmed to expose nonlinear pharmacodynamics. A significant (p = 0.03) contribution of Tempi (but not age) on the value of the slope function was found. There was no consistent effect of age or Tempi on the cyclic component of the integrated model of antipyresis. In addition, a multiple linear relationship of age and Tempi was observed with a large number of the PK, link, and PD variables in those who received IBU. Dose, age, and Tempi interacted with beta in a significant multiple linear relationship with AUC0-infinity. The effects of IBU dose, age, and Tempi are pervasive and cascade down the chain of events leading to the PD response. The etiology of pyresis may create the slope function, the magnitude of which may be partially due to the underlying disease. In some cases, the cyclic function may be explained by temperature regulation. Regardless of their cause, both confound analysis of drug action and make the simple, unmodified Sigmoid Emax effect model less than satisfactory for interpretation of antipyretic drug effects. The influence of Tempi on the magnitude of antipyretic drug response is also a finding with major impact on PD investigations of antipyretic medications. In children receiving IBU, dose and age are also confounders, in addition to Tempi. A multiplicity of covariables must be taken into account when developing appropriate dosing regimens for these antipyretics in febrile children.