1. The effects of adenosine on adenosine 5'-triphosphate (ATP)-evoked dopamine release from rat phaeochromocytoma PC12 cells was investigated to determine whether adenosine exerts a regulatory effect on the ATP-evoked response. Adenosine potentiated ATP (30 microM)-evoked dopamine release in a concentration-dependent manner over a concentration-range of 1 to 100 microM. Adenosine (100 microM) shifted the concentration-dependence of the ATP-evoked response to the left without affecting the maximal response. 2. Aminophylline, a non-selective adenosine receptor antagonist, and CP66713, a selective antagonist at the A2 subclass of adenosine receptors, abolished the adenosine-induced potentiation. Furthermore, 8-cyclopentyltheophylline, a selective antagonist at the adenosine A1 receptor partially inhibited the adenosine-evoked potentiation. CGS22492, a selective A2 receptor agonist, potentiated ATP-evoked dopamine release whereas N6-cyclohexyladenosine (CHA), a selective A1 receptor agonist, had no effect. 3. Pertussis toxin (PTX), a bacterial exotoxin which catalyzes the ADP-ribosylation of guanosine 5'-triphosphate (GTP)-binding proteins (G-proteins), inhibited the adenosine-induced potentiation of dopamine release. Dibutyryl cyclic AMP (db cyclic AMP), an analogue of cyclic AMP, had no effect on the release on the ATP-evoked response. 4. Adenosine potentiated the ATP-evoked rise in intracellular Ca2+ concentration ([Ca]i) in PC12 cells. This potentiation was also observed with CGS 22492 but not with CHA. PTX completely inhibited the adenosine-induced potentiation of the rise in [Ca]i. 5. On the basis of these findings, we suggest that the adenosine-induced potentiation of ATP-evoked dopamine release was due to an increase in [Ca]i in the cells. Although the potentiation is most likely mediated by a subclass of A2 receptors, the subclass may be different from those previously reported since the potentiation was sensitive to PTX and was not reproduced by db cyclic AMP.