Most of the myriad (order of ) interconnected coronary vessels interact nonlinearly with their embedding contracting myocardium. Their dynamic flow can be simulated based on a nonlinear distributive segmental flow model involving highly nonlinear partial differential equations. Such network flow analysis, although of high accuracy, is computationally excessively complex. On the other hand, a corresponding nonlinear lumped analysis is significantly less demanding since it involves ordinary differential equations. This is, however, at the detriment of accuracy. In the present technical report, a nonlinear lumped representation of coronary segmental flow is presented and tested against predictions of the corresponding distributive analysis. The results suggest that under physiological conditions, the proposed lumped model achieves similar accuracy to the distributive one, yet with considerably higher computational speed.