We present a new quantitative framework for analyzing epicardial activation patterns recorded by high-resolution mapping during ventricular fibrillation (VF). Individual activation wavefronts are isolated from one another using an algorithm that groups adjacent active electrogram samples (dV/dt<-0.5V/s). The timing and the interactions between wavefronts are summarized in the form of a directed graph which is then analyzed to extract quantitative measures of the activation pattern, for example, the number of wavefronts, the number of wavefront fractionations and collisions, the mean size and area swept out by wavefronts, and the incidence of reentry. To evaluate this methodology, we analyzed electrically induced VF in 6 pigs during the first 20 s post-induction. The results indicate a recovery of organization consistent with previous findings. The incidence of reentrant circuits was low (5%).