Evaluation of proneness to red light running behavior, because it results from human and road factors, can aid in a proper selection of the sites to be treated and thereby increase the benefit of countermeasures to reduce the red light running (RLR) phenomenon. Starting from the conceptual framework of a model based on potential conflicts analysis, this paper shows that a quantitative evaluation of proneness to red light running behavior can be obtained from both the analysis of the effective operational characteristics of the intersection and the actual number of RLR violations. According to behavioral models referred to in the literature, which emphasize the influence that both human and road factors have on the user's decision-making process at red lights, the proposed approach also accounts for the impact of the local (site) and general (population) characteristics on the phenomenon. Field observations for a case study in an urban area are discussed to illustrate the methodological approach. Results obtained clearly show the sensitivity of the parameter assumed for describing proneness to red light violation to the operational characteristics of the intersections. Moreover, they underline that the common violation rates could cause improper evaluation of the extent of the RLR phenomenon at a specific intersection.