The present study was designed to investigate the role of nitric oxide (NO), guanosine 3',5'-cyclic monophosphate (cGMP), and adenosine 3',5'-cyclic monophosphate (cAMP) in the vasodilator response to beta-adrenoceptor agonists in newborn pigs equipped with a closed cranial window. Dobutamine (10(-8) and 10(-6) M), a beta 1-agonist, produced pial artery dilation that was blunted by N-G-nitro-L-arginine (L-NNA; 10(-6) M), a NO synthase inhibitor (12 +/- 1 vs. 0 +/- 2% and 24 +/- 3 vs. 4 +/- 1% for 10(-8) and 10(-6) M dobutamine, respectively). Dobutamine-induced vasodilation was associated with increased cortical periarachnoid cerebrospinal fluid (CSF) cGMP, and these changes in CSF cGMP were blocked by L-NNA (391 +/- 10 and 615 +/- 36 fmol/ml vs. 307 +/- 3 and 346 +/- 37 fmol/ml for control and 10(-6) M dobutamine before and after L-NNA, respectively). In contrast, dobutamine-associated changes in CSF cAMP were unchanged by L-NNA (1,108 +/- 56 and 2,623 +/- 139 fmol/ml vs. 1,059 +/- 24 and 2,500 +/- 61 fmol/ml for control and 10(-6) M dobutamine before and after L-NNA, respectively). Salbutamol, a beta(2)-agonist, and isoproterenol, a nonselective (b)eta-agonist, elicited similar changes in pial diameter and cyclic nucleotides; vasodilation and changes in CSF cGMP also were similarly inhibited by L-NNA. Forskolin, a direct activator of adenylate cyclase, induced pial dilation that was attenuated by L-NNA (19 +/- 1 and 36 +/- 3% vs. 11 +/- 2 and 29 +/- 3% for 10(-8) and 10(-6) M forskolin before and after L-NNA, respectively). Sodium nitroprusside-induced dilation was associated with increased CSF cGMP but no change in CSF cAMP. These data indicate that cAMP, cGMP, and NO contribute to beta-adrenoceptor-mediated vasodilation. These data further suggest an association between cAMP elevation and the consequent release of NO.