Due to their frequency and magnitude, urban floods affect different regions of the world. For this reason, several methodologies integrate information on hazard (H) and vulnerability (V) using a "Classic" Risk (R) model for risk analysis. However, this combination of variables generally overestimates the risk in places where the frequency of flooding is low. In this work we propose a model that we call Adjusted Risk (AR) that integrates values ​​of urban proximity (p) to bodies of water, as a tool to assess the risk of floods. The comparison between the R and AR models showed a higher efficiency of AR to reproduce the frequency of floods for 210 cities in Veracruz, while R overestimated the level of risk in cities with low frequency of floods. The correlation values ​​associated with the frequency of flood events for a period of 45 years (1970-2015), allow to establish the utility of the AR model to evaluate the risk of urban floods when using different scales of analysis.