Sawynok et al.’have recently summarized their elegant work indicating that the release of adenosine from a population of capsaicin-sensitive primary afferent terminals may mediate the antinociceptive effects of morphine in the dorsal horn of spinal cord. They mention in passing behavioural work that indicates the importance of a similar process at supraspinal levels. However, both neurochemical work in vitro and in vivo and electrophysiological studies in vivo performed ten years ago strongly pointed to the involvement of adenosine in morphine’s effects in brain. this effect was prevented by the application of theophylline at iontophoretic currents that also antagonized responses to adenosine but not GABA’*‘. The implication of this workthat morphine acted by releasing adenosine-was slpported by the demonstration that morphine, at low micromolar concentrations, enhanced the release of adenosine from brain slices induced by several depolarizing agents45 or hypoxia6, or from the surface of the cerebral cortex in anaesthetized rats following systemic administration of the opiate7. All these studies were in turn triggered by the even earlier observation by Ho et al.* that the antinociceptive effect of morphine could be antagonized by xanthines.