The roles played by the cerebral monoamines (dopamine, noradrenaline and serotonin) in stimulation-produced analgesia (SPA) have been investigated in the rat employing the tail flick test. SPA was elicited through bipolar electrodes chronically implanted in the mesencephalic periaqeductal gray matter, an area previously shown to yield potent and reliable analgesic effects. Four approaches were used to alter transmission in monoamine pathways. (1) Depletion of monoamines by administration of tetrabenazine (TBZ), p-chlorophenylalanine (PCPA), alpha-methyl-para-tyrosine (AMPT), or disulfiram. (2) Replacement of depleted monoamine stores by appropiate precursors (5-HTP or L-DOPA) in combination with a peripheral decarboxylase inhibitor. (3) Potentiation of monoamine systems by administration of precursors to previously untreated animals or by administration of a dopamine receptor stimulator, apomorphine. (4) Blockade of catecholamine receptors by haloperidol or of dopamine receptors by pimozide. These four approaches yielded internally consistent results. Depletion of all 3 monoamines (TBZ) led to a powerful inhibition of SPA. Original levels of SPA were restored by injection of either 5-HTP or L-DOPA. Specific depletion of serotonin (PCPA) caused a reduction in SPA, whereas elevation of serotonin levels (5-HTP) caused an increase in SPA. Dopamine receptor blockade (pimozide) decreased SPA, whereas the precursor (L-DOPA) and a dopamine receptor stimulator (apomorphine) increased SPA. On the other hand, selective depletion of noradrenaline (disulfiram) caused an increase in SPA; and at a time when noradrenaline levels are depressed and dopamine levels are elevated (AMPT + L-DOPA), SPA was seen to be particularly enhanced. thus, dopamine and serotonin appear to facilitate SPA, whereas noradrenaline appears to inhibit it. When a general catecholamine receptor blocker (haloperidol) was employed, SPA was diminished, suggesting that the influence of dopamine in SPA is greater than that of noradrenaline. Most of the drugs used in this study significantly altered SPA at doses which left baseline tail flick latency unaffected. It would appear, therefore, that SPA has a neural substrate at least partly independent of that underlying baseline pain responsiveness. Consideration is given to various ascending and descending monoamine system as possible component paths in this neural substrate of SPA. Finally, the present results are discussed in relation to studies by others on the site and mechanism of morphine's analgesic action. Some striking parallels between SPA and morphine analgesia are noted. These suggest the existence of a common pain-inhibitory system in the brain activated by morphine and by focal electrical stimulation.