Research studies concerning the behavioral effects of chlordiazepoxide hydrochloride, diazepam, and meprobamate are analyzed and reviewed. Five major characteristics of studies in the sample are statistically described and related. The drug sensitivity of 43 performance measures is evaluated, most and least promising measures are identified, and performance effects are related to experimental design characteristics. Conflicting evidence, gross inadequacies in the data base and glaring gaps in the literature greatly restrict the clinical implications of the research to date. In view of their widespread use, it should be of concern that there is so little sound and clinically relevant information about what behavioral effects these drugs produce.