The subjective, psychomotor, and physiological effects of three opioid mu-receptor agonists were studied in healthy volunteers using a cumulative-dosing procedure. Sixteen volunteers with no history of drug abuse received i.v. injections of saline (SAL), morphine (MOR), hydromorphone (HM), or meperidine (MEP) in a randomized double-blind crossover design. Subjects received 1 injection/h for the first 4 h, and a 3-h recovery period followed. SAL was injected first during each session, then SAL or increasing doses of each drug were administered every hour for the next 3 h. The absolute doses per injection were MOR: 2.5, 5, and 10 mg/70 kg; HM: 0.33, 0.65, and 1.3 mg/70 kg; and MEP: 17.5, 35, and 70 mg/70 kg. These injections resulted in cumulative doses of MOR: 2.5, 7.5, and 17.5; HM: 0.33, 0.98, and 2.28; and MEP: 17.5, 52.5, and 122.5 mg/70 kg. Subjects completed mood forms and psychomotor tests, and physiological measures were recorded at various times after each injection and during recovery. MEP tended to produce the most intense effects immediately after drug injection, which dissipated rapidly. MOR produced the mildest effects but was associated with unpleasant side effects during recovery and after the session. HM's effects were stronger than MOR's, and the recovery from HM was slower than with MEP. None of the opioids produced consistent effects that are typically associated with abuse liability. Orderly dose-response functions suggested that our cumulative-dosing procedure is an efficient way of determining dose-response functions for multiple opioids within the same subjects within the same study.