1 This study tested the hypothesis that heterogeneity of myocardial adenylyl cyclase activity exists and contributes to the heterogeneity of blood flow in canine myocardium, in which flow varies in different areas at any one moment or in a single area at different times. 2 Tissue adenylyl cyclase activity was measured simultaneously with coronary blood flow in the left ventricle of eight anaesthetized open-chest dogs. Coronary flow was measured using radioactive microspheres under control conditions and after sympathetic stimulation. The ventricle was cut into 15 subepicardial and 15 subendocardial sections in which both flow and adenylyl cyclase (basal and forskolin-stimulated) activity were assayed. 3 Control coronary flow was 73.1+/-19.9 ml min-1 100 g-1 and ansa subclavia nerve stimulation increased flow to 93.2+/-26.1. Baseline adenylyl cyclase activity averaged 51.9+/-38.8 pmol min-1 mg protein-1, while forskolin-stimulated activity was 588.2+/-393.6. Method variability was approximately 10% of the mean, much less than the biological variability of 70%. 4 Regression analysis of control flow vs baseline adenylyl cyclase activity produced the equation: flow=0.25 (Adenylyl Cyclase)+60.6, (r=0.47, P<0.0001). The equation for stimulated flow vs stimulated adenylyl cyclase was: flow=0.03 (Adenylyl Cyclase)+73.1, (r=0.51, P<0.0001). 5 Thus, significant heterogeneity of tissue adenylyl cyclase activity existed in the heart. There was also a direct linear relationship between adenylyl cyclase activity and coronary flow heterogeneity in canine left ventricle.