This study presents a surrogate safety measure for evaluating the rear-end collision risk near recurrent bottleneck area using aggregated traffic data from loop detectors. The attributes of kinematic waves that accompanied rear-end collisions and the traffic conditions at detectors stations spanning the collision location were examined to develop the rear-end collision risk index (RCRI). The logistic regression model was developed using RCRI and standard deviation in occupancy observed at an upstream location. Findings indicated that a unit increase in RCRI results in increasing the odds of rear-end collisions by 32.3%, and an additional unit increase in standard deviation of upstream occupancy increases the odds by 19.8%. The likelihood of rear-end collisions were highest when traffic approaching from upstream location is near capacity while the downstream condition is congested. The proposed model was used to predict rear-end collision risk at the 6-mile study site and compared with the observed traffic collision data from year 2008. Predicted rear-end collision risks were consistent with the observation.