SUMMARY. We analyzed the relationship between diastolic coronary artery pressure and flow in the canine coronary bed, using an electrical analog model of the coronary circulation based on the theory of critical closure. The model contains a voltage-dependent nonlinear resistance and capacitance. The behavior of the resistive element was described using experimental diastolic pressure-flow curves obtained in the absence of compliance effects. Compliance free zero flow pressure intercepts (Pf 0 ) exceeded coronary venous pressure (P v ) by 2-to 5-fold and were related to initial diastolic coronary artery pressure P a (0) and flow F(0), and P v by: P f0 = 14.3 [(P a (0) -P v )/F(0)] + P V + 4.0 (r = 0.93). When coronary artery pressure was suddently lowered to values less than or equal to the compliance-free P ra , diastolic flow abruptly decreased and, after a transient reversal, remained at zero for up to 8 seconds. In the model, zero flow pressure represents critical closing pressure and the resistance regulating flow is the difference between coronary artery and venous pressure divided by flow. Theoretically predicted pressure-flow curves were in good agreement with existing experimental data, including the effects of elevating coronary venous pressure on zero flow pressure. Differences between compliance-free pressure-flow curves and those obtained with pressure gradually decreasing were explained by a coronary arterial compliance whose magnitude varies inversely with pressure and is dependent on vasomotor tone. In conclusion, the results of this study demonstrate the existence of a diastolic pressure gradient across the canine coronary bed at zero flow which is dependent on coronary vasomotor tone. A theoretical model of the coronary circulation based on the concept of critical closure describes the observed relationship between diastolic coronary artery pressure and flow during various experimental conditions. (Circ Res 55: 215-226, 1984)