The safety and efficacy of lorazepam and hydroxyzine, administered intramuscularly for surgical premedication, were compared in a double-blind placebo-controlled study. Ninety patients were given either 0.05 mg/kg of lorazepam (maximum 4 mg), 1.5 mg/kg of hydroxyzine (maximum, 100 mg), or 1 ml of normal saline before surgery. By 60 min after injection, patients given lorazepam or hydroxyzine were significantly (P less than 0.05) more sedated than the placebo-treated patients. The sedative effect of lorazepam lasted longer than that of hydroxyzine. Baseline anxiety scores were low; thus differences in the anxiolytic effect of the test drugs were not detected. Lorazepam-treated patients were consistently less able to recall postinjection events than patients treated with hydroxyzine or placebo. The frequency of injection-site reactions was similar in the three treatment groups. Premedication with lorazepam was considered satisfactory by a higher percentage of patients (77%) than was that with hydroxyzine (62%) or placebo (34%); only lorazepam was considered significantly (P less than 0.005) superior to placebo. The advantages of a longer duration of sedative effect, a greater effect on recall, and higher patient acceptance favor the use of lorazepam over hydroxyzine for surgical premedication.