Implementation of regulations to control the prescribing of benzodiazepines in New York State in 1989 resulted in a 55% decrease in benzodiazepine prescribing, with a concomitant increase in the rates of prescribing older sedative-hypnotic compounds such as butabarbital (30% increase) and meprobamate (125% increase). In a double-blind, crossover, placebo-controlled study, we compared the behavioral and pharmacological effects of triazolam, meprobamate, and butabarbital in 14 recreational drug users. Placebo and three doses each of triazolam, meprobamate, and butabarbital were administered to each subject in a random order. Objective tests (motor performance, concentration) and subjective response questionnaires measured drug effects. Triazolam, meprobamate, and butabarbital showed comparable negative dose-response slopes on the objective measures. On the basis of these objective data, equivalent doses for the three compounds were determined to be as follows: 0.5 mg triazolam = 2,400 mg meprobamate = 400 mg butabarbital. Subjective effects data on equivalent doses show that butabarbital produced the highest peak score on Cole/ARCI Abuse Potential, ARCI Pentobarbital Chlorpromazine Alcohol Group (PCAG), and "drug strength" scales. Triazolam and butabarbital produced equivalent results on ARCI Morphine Benzedrine Group (MBG), Cole/ARCI Euphoria, and "drug liking" scales. Meprobamate was indistinguishable from placebo on euphoria and abuse potential scales. Behavioral economics analysis indicated a price crossover point two times higher for butabarbital (400 mg) than for any other drug condition. These data indicate a comparative abuse liability of butabarbital > triazolam > or = meprobamate, suggesting that the prescribing restrictions on benzodiazepines had little net benefit on abuse risk in the population and may have increased the risk of overdose morbidity and mortality.