To investigate whether ATP participates in spinal nociceptive transmission, effects of intrathecally applied P2-purinoceptor antagonists and agonists in the tail-flick and the formalin test were studied in rats. In the tail-flick assay, the P2 antagonists suramin (12-120 micrograms), Evans blue (0.1-10 micrograms), Trypan blue (1-30 micrograms) and Reactive blue 2 (1-30 micrograms) but not pyridoxalphosphate-6-azophenyl-2',4'-disulfonic acid (PPADS; 0.03-30 micrograms) caused moderate antinociception up to a doubling of the response latency. In contrast, the P2 agonists alpha,beta-methylene ATP (alpha,beta-mATP, 0.3-30 micrograms) and 2-methylthio-ATP (3-30 micrograms) decreased the tail-flick latency by up to about 50%. When co-injected with alpha,beta-mATP, suramin (120 micrograms) or Evans blue (10 micrograms) prevented the effect of alpha,beta-mATP 3 micrograms but not of alpha,beta-mATP 30 micrograms. In the formalin test, pretreatment with suramin (3-90 micrograms) 60 min prior to testing caused significant antinociception by decreasing the weighted pain intensity score by up to about 80%. alpha,beta-mATP (30 micrograms), applied 30 min prior to testing, was without effect. The results indicate that endogenous ATP, acting through P2-purinoceptors, may contribute to nociceptive information processing in the spinal cord.