Perfused rat hearts were used to study the effects of acidosis on coronary tone. When pH was decreased, over the range pH 7.4 to pH 6.2, by reducing perfusate bicarbonate levels, under constant flow conditions, there was a transient decrease in coronary perfusion pressure (CPP), followed by a sustained acidosis-dependent increase in CPP, which reversed when pH was returned to pH 7.4. This increase in CPP was seen at perfusion rates of 5, 10, and 20 ml/min(-1). When using constant pressure perfusion acidosis reduced coronary flow. In a HEPES-buffered bicarbonate-free solution, acidosis did not cause a transient fall in CPP but it did produce a sustained increase in CPP. Addition of ammonium chloride (10 mM) reduced CPP, while washout of ammonium chloride increased CPP. The acidosis-induced increase in CPP was not affected by indomethacin, nitro-L-arginine, the nonselective adenosine receptor antagonist, 8-phenyl theophylline, or the thromboxane receptor antagonist, ZD 1542. The acidosis-induced increase in CPP was independent of the myocardial depressant effects of acidosis, but was attenuated by three different L-type calcium channel blockers. These results demonstrate that the coronary circulation of the rat constricts in response to acidosis. Experiments performed with L-type calcium channel blockers, and the calcium channel activator BAY K8644, suggest that constriction occurs via activation of L-type calcium channels. This would not be expected on the basis of electrophysiological studies, which have shown an inhibition of L-type calcium channels by acidosis.