Traffic systems are man-made systems in which the human component as a complex factor plays a significant part. (1) The realistic modeling of them is extremely complicated. The aim of the paper is to present a traffic behavior model of inhomogeneous driver-populations and a corresponding estimation method. The traffic behavior model is developed for describing traffic flow in a signalized intersection. It works with inhomogeneous driver populations; different type of drivers may make different decisions in the same traffic situation. This way various driver types can be defined like aggressive, normal or conservative etc. drivers. The model focuses on the movement and the behavior of the inhomogeneous driver-populations approaching a signalized intersection, passing the amber signal light. The developed estimation method is intended to define the distribution of different driver types. This theory can allow an extensive and more complex traffic analysis on a more sensitive scale. The paper details the steps of the estimation process and the validation and introduces the results of the application of the estimation method with real experiments.