The pharmacodynamic interaction between midazolam and its active metabolite alpha-OH-midazolam was investigated to evaluate whether estimates of relevant pharmacodynamic parameters are possible after administration of a mixture of the two. Rats were administered 10 mg/kg of midazolam, 15 mg/kg of alpha-OH-midazolam, or a combination of 3.6 mg/kg of midazolam and 35 mg/kg of alpha-OH-midazolam. Increase in the 11.5- to 30-Hz frequency band of the electroencephalogram was used as the pharmacodynamic endpoint. The pharmacodynamics of midazolam and alpha-OH-midazolam after combined administration were first analyzed according to an empirical and a competitive interaction model to evaluate each model's capability in retrieving the pharmacodynamic estimates of both compounds. Both models failed to accurately estimate the true pharmacodynamic estimates of midazolam and alpha-OH-midazolam. The pharmacodynamic interaction was subsequently analyzed according to a new mechanism-based model. This approach is based on classical receptor theory and allows estimation of the in vivo estimated receptor affinity and intrinsic in vivo drug efficacy. The relationship between stimulus and effect is characterized by a monotonically increasing function f, which is assumed to be identical for midazolam and alpha-OH-midazolam. The pharmacodynamic interaction is characterized by the classical equation for the competition between two substrates for a common receptor site. This mechanism-based interaction model was able to estimate the pharmacodynamic parameters of both midazolam and alpha-OH-midazolam with high accuracy. It is concluded that pharmacodynamic parameters of single drugs can be estimated after a combined administration when a mechanistically valid interaction model is applied.