Rat dorsomedial medullary brain segments containing primarily nucleus tractus solitarius (NTS) were employed for slice superfusion studies of electrically evoked [H-3]serotonin ([H-3]5-HT) release. Individual slices loaded with [H-3]5-HT were stimulated two times, S-1 and S-2, at 3 Hz, 25 mA, 2 ms pulses for 1 min. Control NTS slices had a S-2/S-1 ratio of 0.94(+/-0.02). Superfusion of tissue slices with 0.1 nM to 100 nM 2-p-(2-carboxyethyl)-phenethylamino-5'-N-ethylcarboramidoadenosine (CGS 21680), a selective adenosine A(2a) receptor agonist, for 5 min-prior to the S-2 stimulus produced a significant concentration-dependent increase in the S-2/S-1 fractional release ratio which was maximal (37.2% increase, P < 0.01) at 1.0 nM. However, superfusion of tissue slices with CGS 21680 over the same concentration range for 20 min prior to the S-2 stimulus did not significantly alter the S-2/S-1 ratio from control release ratios. The augmented release of [H-3]5-HT mediated by 1.0 nM CGS 21680 with 5 min tissue exposure was abolished by 1.0 nM 9-chloro-2-(2-furanyl)-5,6-dihydro-[1,2,4]-triazolo[1,5-c]quinazolin-5-imine (CGS 15943) as well as by 100 nM 8-(3-chlorostyryl)caffeine (CSC), both A(2a) receptor antagonists, but not by 1.0 nM 8-cyclopentyl-1,3,-dipropylxanthine (DPCPX), the A(1) receptor antagonist. These results indicate that CGS 21680 augmented the evoked release of [H-3]5-HT in the NTS by way of activation of presynaptic adenosine A(2a) receptors. It was also apparent that this population of adenosine A(2a) receptors in the NTS desensitized within 20 min since the augmenting action of CGS 21680 on evoked transmitter release was not evident at the longer time interval.