Triazolam was used to study the plasma concentration-effect relationship of a benzodiazepine because it has a very short plasma t1/2. A standard hypnotic dose of 0.25 mg was given by mouth to six healthy subjects, and blood samples were drawn when the subjects had to perform a battery of psychologic tests. Only the digit-symbol substitution test, the card-sorting test according to numbers, and the visual analog scale (energetic-lethargic) gave significant results. Analysis of the concentration-effect relationship in individuals indicated a wide scatter of the data. Mean values revealed a trend for a learning effect in the card-sorting test. The results are consistent with the hypothesis that triazolam is well suited for a study of concentration-effect relationships, but better psychologic tests would be desirable.