This research identified infrastructural and operational factors that influenced the most common type of car–tram collision: cars making opposing turns in front of trams. Few studies have analyzed influences on car–tram collisions quantitatively, but none have explored predictor factors for opposing-turn crashes—a research gap addressed with this paper. The two largest Swiss tram networks, Basel and Zurich, were used for the analysis. A point-based research approach was chosen: all locations within a tram network at which a car could turn left (an opposing turn where traffic drives on the right) in front of a tram were identified. For each of these points, data on dependent and predictor variables were collected. This data set was analyzed with Poisson, negative binomial, and zero-inflated negative binomial regression models. The number of left-turning car–tram collisions was used as the dependent variable, while predictors were derived from a literature review; models were fitted by using all predictors and with forward variable selection by means of Akaike’s information criterion. Traffic volumes (cars and trams), tram speed, and dedicated left-turn lanes were found to be significantly associated with a higher frequency of car–tram collisions, whereas turning left to access a service rather than a road, left-turn restrictions, proximity to a tram stop, and perpendicular turning angles were significantly associated with a lower frequency of left-turning car–tram collisions. On the basis of these results, left turns across tramways should be restricted for cars. Remaining conflict points should be located close to tram stops, have limited tram speed, and feature perpendicular turning angles.