We studied the effects of intracoronary injection of nicotine and acetylcholine on coronary blood flow in nine conscious calves chronically instrumented to measure coronary blood flow, left ventricular (LV) and mean arterial pressure, LV dP/dt, and heart rate. Nicotine (5 micrograms/kg i.c.) elicited a biphasic response in coronary blood flow consisting of an initial vasoconstriction (phase 1; blood flow fell by 52 +/- 5.4% from a baseline of 66 +/- 7.5 ml/min) followed by vasodilation (phase 2, blood flow rose 119 +/- 12.7% above baseline). The change in coronary blood flow with nicotine was not associated with changes in LV systolic pressure, mean arterial pressure, or heart rate. The change in coronary blood flow was unaffected by combined alpha- and beta-adrenoceptor blockade with prazosin, rauwolscine, and propranolol but was abolished by either muscarinic blockade with atropine or ganglionic blockade with hexamethonium. Acetylcholine (0.5 microgram/kg i.c.), without affecting mean arterial pressure, elicited changes in coronary blood flow similar to those observed with nicotine, producing an initial phase of coronary vasoconstriction (blood flow fell by 71 +/- 4.9%) followed by vasodilation (blood flow rose by 228 +/- 20.7%). Both phases of the response to acetylcholine were abolished by muscarinic blockade but were unaffected by ganglionic blockade. When nicotine was injected into the left circumflex coronary artery, no change in blood flow was observed in the left anterior descending coronary artery, indicating the lack of involvement of global reflex pathways. These results suggest that nicotine locally stimulates parasympathetic nerves, which constrict the coronary circulation via a muscarinic mechanism.