Background and purpose:This study establishes a pharmacokinetic/pharmacodynamic (PK/PD) model to describe the time course and in vivo mechanisms of action of the antinociceptive effects of lumiracoxib, evaluated by the thermal hyperalgesia test in rats. Experimental approach: Female Wistar fasted rats were injected s.c. with saline or carrageenan in the right hind paw, followed by either 0, 1, 3, 10 or 30 mg·kg -1 of oral lumiracoxib at the time of carrageenan injection (experiment I), or 0, 10 or 30 mg·kg -1 oral lumiracoxib at 4 h after carrageenan injection (experiment II). Antihyperalgesic responses were measured as latency time (LT) to a thermal stimulus. PK/PD modelling of the antinociceptive response was performed using the population approach with NONMEM VI. Results: A two-compartment model described the plasma disposition. A first-order model, including lag time and decreased relative bioavailability as a function of the dose, described the absorption process. The response model was: LT = LT0/(1 + MED). LT0 is the baseline response, and MED represents the level of inflammatory mediators. The time course of MED was assumed to be equivalent to the predicted profile of COX-2 activity and was modelled according to an indirect response model with a time variant synthesis rate. Drug effects were described as a reversible inhibition of the COX-2 activity. The in vivo estimate of the dissociation equilibrium constant of the COX-2-lumiracoxib complex was 0.24 mg·mL -1. Conclusions: The model developed appropriately described the time course of pharmacological responses to lumiracoxib, in terms of its mechanism of action and pharmacokinetics.