Glibenclamide, iberiotoxin, and apamin (blockers of ATP-sensitive, large-conductance, and small-conductance Ca(2+)-activated potassium channels, respectively) were infused into the diaphragmatic vasculature of anesthetized dogs to assess the contribution of these channels in the regulation of basal tone and the response to brief occlusions of the left phrenic artery (reactive hyperemia). Baseline phrenic flow (Qphr), peak postocclusive flow, and reactive hyperemia duration in response to 10-, 30-, 60-, and 120-s arterial occlusions were measured before (control) and after the infusion of K+ channel blockers in three groups of animals. Glibenclamide at 5 x 10(-6), 1 x 10(-5), and 8 x 10(-5) M increased baseline phrenic resistance to 140, 204, and 192% of control values, respectively. Peak postocclusive Qphr and duration of hyperemia in response to all occlusion durations were significantly attenuated after glibenclamide infusion. Iberiotoxin infusion at 1 x 10(-8), 3 x 10(-8), and 1 x 10(-7) M increased phrenic resistance to 141, 133, and 146% of control values, respectively. By comparison, baseline phrenic resistance rose to 159 and 145% of control in response to 1 x 10(-7) and 1 x 10(-6) M apamin, respectively. Iberiotoxin and apamin reduced peak postocclusive flow and duration of hyperemia only in response to 10- and 30-s occlusions. We infused K+ channel blockers along with lemakalim into the diaphragm during constant flow perfusion in separate groups of animals. When infused alone, lemakalim reduced phrenic resistance by 60-70%.(ABSTRACT TRUNCATED AT 250 WORDS)