The bioavailabilities of generic and reference promethazine 50 mg rectal suppositories were compared with that of 50 mg reference oral solution (24 subjects), and all three treatments were compared with a 50 mg reference i.m. injection (six subjects). Plasma samples were assayed by an HPLC method with triflupromazine as the internal standard. Both suppositories produced lower peak plasma concentrations (Cmax) and longer times to peak concentration (Tmax) than did the oral solution. There were no significant differences in the mean area under the plasma concentration-time curves (AUC) from 0 to 24 h among the three treatments. The Cmax of the i.m. injection was significantly higher than the other three treatments, while the Tmax of the injection was significantly shorter than the reference suppository only. The mean AUC of the injection was significantly greater than the AUCs of the other three treatments. Rectal suppositories of promethazine are more slowly absorbed than oral solutions or i.m. injections; rectal suppositories and oral solutions are less bioavailable than i.m. injections. Diminished systemic bioavailability may result from extensive first-pass hepatic metabolism that occurs after both oral and rectal dosing. There is a high degree of intersubject variation in the bioavailability of promethazine rectal suppositories and oral solutions.