The maintenance of constant cerebral blood flow (CBF) as arterial blood pressure is reduced, commonly referred to as CBF-pressure autoregulation, is typically characterized by a plateau until the vasodilatory capacity is exhausted at the lower limit, after which flow falls linearly with pressure. We investigated the effect of cortical, as opposed to systemic, nitric oxide synthase (NOS) inhibition on the lower limit of CBF-pressure autoregulation. Forty-four Sprague-Dawley rats were anesthetized with halothane and N2O in O2. With a closed cranial window placed the previous day in a ventilated and physiologically stable preparation, we determined the CBF using laser-Doppler flowmetry. Animals with low reactivity to inhaled CO2 and suffused ADP or ACh were excluded. Five arterial pressures from 100 to 40 mmHg were obtained with controlled hemorrhagic hypotension under cortical suffusion with artificial cerebrospinal fluid (aCSF) and then again after suffusion for 35 (n = 5) and 105 min (n = 10) with aCSF, 10(-3) M Nomega-nitro-L-arginine (L-NNA; n = 12), or 10(-3) M Nomega-nitro-D-arginine (D-NNA; n = 5). An additional group (n = 7) was studied after a 105-min suffusion of L-NNA followed by a single blood withdrawal procedure. The lower limit of autoregulation was identified visually by four blinded reviewers as a change in the slope of the five-point plot of CBF vs. mean arterial blood pressure. The lower limit of 90 +/- 4.3 mmHg after 105 min of 1 mM L-NNA suffusion was increased compared with the value in the time-control group of 75 +/- 5.3 mmHg (P < 0.01; ANOVA) and the initial value of 67 +/- 3.7 mmHg (P < 0.001). The lower limit of 84 +/- 5.9 mmHg in seven animals with 105 min of suffusion of 1 mM L-NNA without previous blood withdrawal was significantly increased (P < 0.01) in comparison with 70 +/- 1.9 mmHg from those with just aCSF suffusion (n = 37). No changes in lower limit for the other agents or conditions, including 105 or 35 min of aCSF or 35 min of L-NNA suffusion, were detected. The lack of effect on the lower limit with D-NNA suffusion suggests an enzymatic mechanism, and the lengthy L-NNA exposure of 105 min, but not 35 min, suggests inhibition of a diffusionally distant NOS source that mediates autoregulation. Thus cortical suffusion of L-NNA raises the lower limit of autoregulation, strongly suggesting that nitric oxide is at least one of the vasodilators active during hypotension as arterial pressure is reduced from normal.