We have examined interactions between mid-azolam, propofol and alfentanil using two end-points of light sedation (hypnosis) and anaesthesia. Quantal dose-response curves were determined in 400 female patients for the drugs individually and in combination. At the hypnotic end-point, interactions were analysed by fitting the data to a mathematical model where the response depended on the doses of the three drugs with additional terms included to describe non-additive interactions of the various combinations of the three drugs. There were significant interactions for hypnosis; the decrease in expected ED50 for the various combinations were: midazolam-propofol = 37%, midazolam-alfentanil = 46% propofol-alfentanil = 20% midazolam-propofol-alfentanil = 42% Whilst all responses to the two-drug combinations were synergistic, the three-drug combination led to a response that was less than that expected from the effects of the individual agents and their two drug interactions. For anaesthesia, dose-related effects could not be demonstrated for midazolam or alfentanil when used alone. The decrease in ED50 of propofol in the presence of the other compounds was propofol-midazolam = 52%, propofol-alfentanil = 73%, propofol-midazolam-alfentanil = 82%. When comparing the different combinations, the responses varied markedly at each end-point assessed and could not be predicted from the responses of the individual agents.