OBJECTIVE Cerebral arteriolar dilation to N-methyl-D-aspartate (NMDA) is a neuronally mediated multistep process that is sensitive to cerebral hypoxia and ischemia (H/I). We tested the hypothesis that topical pretreatment with the selective potassium channel agonists NS1619 and aprikalim preserves the vascular response to NMDA after consecutive H/I. METHODS Pial arteriolar diameters were measured in anesthetized piglets with the use of a closed cranial window and intravital microscopy. Arteriolar responses to NMDA (10(-5), 5 x 10(-5), and 10(-4) mol/L) were recorded before and 1 hour after 10 minutes of hypoxia (8.5% O2 in N2) plus 10 minutes of ischemia (H/I). Ischemia was induced by increasing intracranial pressure. Subgroups were topically pretreated with 10(-5) mol/L NS1619, 10(-6) mol/L aprikalim, 10(-6) mol/L calcitonin gene-related peptide (CGRP), or 10(-5) mol/L papaverine. We also examined the effects of H/I on vascular responses to kainate (10(-4) mol/L) to assess specificity of neuronal injury. RESULTS Arteriolar responses to NMDA were significantly attenuated after H/I. Baseline compared with post-H/I arteriolar diameters were 9+/-4% versus 3+/-2% at 10(-5) mol/L, 22+/-4% versus 4+/-2% at 5 x 10(-5) mol/L, and 33+/-4% versus 7+/-2% at 10(-4) mol/L (mean+/-SE; all P<.05, n=7). Pretreatment with NS1619 and aprikalim preserved the arteriolar responses to NMDA after H/I. For NS1619 (n=6), values were as follows: 9+/-2% versus 6+/-4% at 10(-5) mol/L, 19+/-6% versus 21+/-5% at 5 x 10(-5) mol/L, and 35+/-3% versus 31+/-5% at 10(-4) mol/L. For aprikalim (n=7), values were as follows: 6+/-2% versus 8+/-2% at 10(-5) mol/L, 22+/-6% versus 15+/-3% at 5 x 10(-5) mol/L, and 41+/-5% versus 32+/-6% at 10(-4) mol/L. In contrast, piglets pretreated with CGRP (n=6) or papaverine (n=5) showed no preservation of the vascular response to NMDA after H/I, although these compounds dilated the arterioles to an extent similar to that with NS1619/aprikalim. Kainate-induced arteriolar dilation (n=6) was largely preserved after H/I compared with preischemic responses. CONCLUSIONS (1) Vascular responses of cerebral arterioles to NMDA after H/I are preserved by pretreatment with NS1619 or aprikalim, indicating a neuroprotective effect. (2) CGRP and papaverine do not preserve the vascular response to NMDA despite causing vasodilation similar to that with NS1619 or aprikalim. This suggests that activation of potassium channels on neurons accounts for the protective effect of potassium channel agonists. (3) Preserved arteriolar dilation to kainate suggests largely intact functioning of neuronal nitric oxide synthase after H/I.