In this paper, an attempt to identify and characterize minor road traffic behaviorat unsignalized urban intersections in a city situated in a developingcountry is described. Information about the particular traffic behavior was collected using video cameras placed at 12 unsignalized intersections connecting minor to major roadways in Tehran. Stop signs and/or blinking red signals provided the stopping message at the minor approaches of the selected intersections.
Relevant information about the behavior of more than 2400 vehicles was extracted from the video displays and the database records consisted of 31 variables reflecting driver, pedestrian, passenger, vehicle and intersection characteristics. Four key driver behavior characteristics, including observation of the stop message,departure distance and time from the stopping position and stopping distance from the pavement markings stopping line, were studied and evaluated. Univariateand multivariate statistical analyses of the database were carried out. Inappropriate and proper drivingbehavior, as related to 4 keycharacteristics, wereidentified and modeled. The applied mode Jingtechniques consisted of regression analysis, artificial neural network modeling and discriminant analysis. Although the study findings are based on a rather limited database and are location specific, the same methodology can be applied to any unsignalizedinter section.