Repeated administration of opioid drugs results in tolerance, a lessening of the drug's effect. There is pre-clinical evidence suggesting a conditioning component to drug tolerance. In the present study, six former opiate dependent subjects received i.v. opiate either by un-signalled infusion or by signalled self-injection and the effects were compared with those of saline under double-blind conditions. The subjects' pre-injection rituals constitute a signal which reliably predict the appearance of the opiate. These rituals produced drug-opposite physiological responses which resulted in an attenuation of the effects of the drug. Thus, tolerance was observed when the subjects injected the opiate, but not when the same dose was received by un-signaled intravenous infusion. These results are consistent with a conditioning explanation for the observed drug tolerance.