This study investigates the traffic accidents between the years 2015 and 2019 due to the imprudence of the driver in the conurbation Viña del Mar y Valparaíso. Thus, critical segments of the road network were identified with the spatial analysis with the KDE+ tool. As opposed to the Kernel density estimation that generates a surface in two dimensions, KDE+ determines the statistically significant critical segments in one dimension along the road network. Results identified the road segments with high collective risk in each city. When comparing the results for both cities, Viña del Mar presents a higher collective risk than Valparaiso in the number of critical segments and the value of this risk. In addition, the majority of the critical segments are located in the downtown area of both cities. Through the results of this study, authorities may assign resources to implement new measures in the most critical segments for imprudent drivers, and thus, reduce the traffic accident rate in the conurbation.