To compare the contribution of nitric oxide (NO) to the buffering of short-term and circadian fluctuations of arterial blood pressure with that of the baroreceptor reflex, conscious foxhounds were subjected to continuous 24-h blood pressure recordings. A pressure transducer was placed into the lumen of the abdominal aorta. Telemetry recordings were done under control conditions, following blockade of NO formation by intravenous bolus injection of N-G-nitro-L-arginine (L-NNA; 16.5 +/- 2 mg/kg body wt) and after total sinoaortic and cardiopulmonary denervation in five dogs each. L-NNA produced a sustained elevation of mean arterial pressure (MAP; 137.2 +/- 6.4 mmHg vs. control, 112.9 +/- 3.7 mmHg). After denervation, no significant increase of MAP was found (113.5 +/- 4.1 mmHg), but the standard deviation of the MAP histogram was significantly greater (22.5 +/- 3.1 vs. 10.6 +/- 0.9 mmHg, P < 0.05). Sequential spectral analysis showed that total power between 0 and 0.5 Hz was elevated more than twofold after L-NNA (P < 0.05). This was due primarily to increased power in the range above 0.1 Hz. After denervation, total power increased about threefold (P < 0.05), almost exclusively occurring below 0.04 Hz. Power in the range above 0.2 Hz was diminished, although not significantly. It is concluded that in the conscious dog, NO, as well as the baroreceptor reflex, is an effective blood pressure buffer. NO is most effective above 0.1 Hz, whereas the baroreceptors primarily buffer fluctuations slower than 0.04 Hz.