A hospital-pharmacy-based program to decrease inappropriate prescribing of triazolam is described, and the program's effect on triazolam prescribing patterns for both hospitalized and nonhospitalized patients is reported. Pharmacists developed criteria for the triazolam "target drug program" in cooperation with the pharmacy and therapeutics committee. Pharmacists contacted prescribers when bedtime triazolam doses were ordered that (1) exceeded 0.125 mg, (2) were for patients greater than or equal to 60 years of age, and (3) were not to be given immediately before a surgical or diagnostic procedure. The pharmacists recommended to the prescribers that triazolam be started at a dosage of 0.125 mg in elderly or debilitated patients and that the dosage be adjusted according to the patient's response, with a repeat dose given if appropriate. Data were collected on triazolam use in the hospital before and after the pharmacist intervention program began. Data also were collected on sales of triazolam to community pharmacies near the hospital. The number of inpatient triazolam orders meeting the criteria for intervention decreased. The percentage of inpatient interventions that resulted in a dosage change also decreased, probably because prescribers' orders for higher triazolam dosages were more likely to be appropriate after the target drug program. Community pharmacies purchased more 0.125-mg triazolam tablets and fewer tablets of greater strengths after the target drug program. Intervention by hospital pharmacists caused a change in the inpatient prescribing pattern for triazolam and appeared to cause a parallel change in the community surrounding the hospital.