This investigation developed strategies along which the anticonvulsant effect of oxazepam in the rat could be pharmacokinetically modeled. After determination of the pharmacokinetics of oxazepam, which could be described with a two-compartment model (half-lives of distribution and elimination 6 and 52 min, respectively), the drug was administered iv to groups of animals to achieve a serum concentration range of 0.1-2.5 mg/L at 10, 45, and 120 min after administration. At these time points pentylenetetrazol (PTZ) was infused slowly until the first myoclonic jerk occurred. The anticonvulsant response, expressed as the elevation of the serum or brain threshold concentration of PTZ, was modeled versus the serum (both total and free) and brain oxazepam concentration, according to the sigmoid Emax model. The total serum and brain oxazepam EC50 values are about 0.5 mg/L and 1.1 mg/kg, respectively, and Emax 120 mg/L PTZ. No marked differences in pharmacodynamic parameters between the three time groups were found, which indicates that serum and brain are pharmacokinetically indistinguishable from the effect compartment, that there is no (inter)activity of oxazepam metabolites and absence of development of acute tolerance during the investigated time frame. An interfering role of metabolites was also excluded by a direct radioreceptor assay of oxazepam, yielding very similar results as the specific chromatographic assay. It is concluded that the conception-anticonvulsant effect relationship of oxazepam can satisfactorily be described by the sigmoid Emax model, when utilizing the employed experimental strategies.