Abstarct: signalized intersections, despite being a small part of all roads, are considered to Accident prone. scheduled intersections drivers start time when the yellow light and the proximity to the dilemma zone say, close to the intersection Usually has two choices, continue to move through the intersection or behind the line stops pedestrian. Among the drivers who have to cross the intersection, They number empty the of intersection but others fail, This interferes with the movement of other vehicles or even collides with them. This issue has caused the study of the dilemma zone to be of special importance. Modeling stop passing drivers an important role in prevention systems dilemma zone to avoid being caught driving in the dilemma zone. In this thesis, in order to model the stopping and passing of drivers using data from three intersections in the city of Semnan, Six model groups were examined by SPSS and NLOGIT software. The results show that the vehicle speed at the beginning of the yellow phase has a negative impact on driver behavior Because with increasing speed, the time required for the driver in the decision makings declined. It was also found that time to stop line (TTS) is a better indicator compared to distance to stop line (DTS). When the yellow light is on, the distance from the car to the intersection should be shorter, Increases the possibility of drivers crossing the yellow light. It should be noted that in the end, to evaluate and compare the models, their validation is done, The logistic regression model baxsed on time to the stop line with an accurate estimation percentage of 99.43 % has been introduced as the best model.