The pharmacokinetic (PK) and pharmacodynamic (PD) characteristics of the neuromuscular blocking agent mivacurium were evaluated separately in two groups of rats receiving 0.6 mg kg-1 of mivacurium in a 2.5-min intravenous continuous (iv) infusion. The PK parameters for mivacurium were determined in the first group. A two-compartment model describes the kinetics of mivacurium in plasma. The estimates of the apparent volume of distribution at steady-state and plasma clearance [mean(SE)] were 650 (123) mL kg-1 and 9.9 (0.75) mL min-1 kg-1, respectively. In the second group, the evoked tibialis anterior muscle tension was monitored. The PK parameters derived from the first group were used to compute mivacurium plasma concentrations (C) at the times the PD measurements were recorded in the second group. The concentration-neuromuscular effect [% depression of initial twitch tension (E)] relationship was analyzed by two approaches. (1) The relationship of estimated effect site concentrations versus E; a sigmoidal Emax model described the effect compartment concentrations versus E relationship. The estimate [mean(SE)] of Cess50 (steady-state plasma concentration eliciting half of maximum E) was 0.65 (0.01) microgram mL-1. The value [mean-(SE)] of Keo (rate constant of equilibration between plasma and effect site) was estimated at 0.32 (0.03) min-1. (2) The relationship of descending limb C versus E; a sigmoidal Emax model described such relationship. The estimate [mean(SE)] of C50 (post-infusion C eliciting half of maximum E) was 0.57(0.03) microgram mL-1. The PD properties of mivacurium were also evaluated in another two groups of animals receiving either 5- or 10-min continuous iv infusion; PK and PD parameters obtained from the 2.5-min infusion experiments were used to predict the time course of E in the groups receiving 0.6 mg kg-1 of mivacurium in 5- and 10-min infusions; simulations using the estimated parameters adequately describe the time course of E in those groups. The effect of mivacurium on the mean arterial blood pressure (MAP) was also investigated; a 10% nonsignificant decrease (p > 0.05) in MAP was found in all groups.