Acetylcholine (Ach) could serve as a selective renal medullary vasodilator in studies of the mechanism of arterial pressure regulation; however, effects of intramedullary Ach infusion were disparate. In anesthetized rats, the total renal blood flow (RBF) was measured by renal artery probe, and local perfusion of the cortex (CBF), outer medulla (OMBF) and inner medulla (IMBF) as laser-Doppler (l-D) flux. Renal artery infusion of Ach (60-150 microg/kg/h) significantly increased RBF by 17% and l-D parameters by 7-14%, without affecting arterial blood pressure (BP); the responses were prevented by inhibition of nitric oxide (NO) synthesis with N(omega)-nitro-L-arginine methyl ester (L-NAME). Intramedullary Ach (180 microg/kg/h) significantly increased OMBF by 9% and IMBF by 7% but also RBF and CBF (both 9%). Carbamylcholine (Cch, 15 or 30-60 microg/kg/h), a stable Ach analog, increased CBF, OMBF, and IMBF by 5-8%; there was no dose dependency and, as with Ach, no preferential effect on medullary perfusion. Intramedullary infusion of prostaglandin E(2) (PGE2) (15 microg/kg/h), and S-nitroso-N-acetyl-D,L penicillamine (SNAP), an NO donor (15-30 mg/kg/h), significantly and substantially increased OMBF and IMBF only. Ach increased perfusion of all kidney zones by an NO-dependent mechanism. Infusion of Ach or Cch into the renal medullary interstitium failed to affect preferentially the medullary perfusion, in contrast to the well-demonstrable selective effects of PGE2 and SNAP. The reason was probably the Ach's dual opposed action, vasoconstrictor and vasorelaxant, on the intrarenal vasculature.