In this research, generalized linear models are used for modeling the trip modal split of Tehran residents. Characteristics of trip and traveler are used for calibrating the model. 434 questionnaires are distributed between random samples of students of Tehran. Selected students were asked to fill one questionnaire form about trips of their family and the other questionnaire about trips of their neighbor. The gathered data of questionnaires are analyzed and using a log linear model, effect of trip destination, trip goal, and trip time on the modal split is studied. Sample size is determined as it can produce the necessary information with sufficient accuracy. It is emphasized that the sample size is dependant to the information and accuracy which is needed for decision making. It has been shown that a sample size of 400 families in Tehran is enough for determining the trip goals. For determining the trip destinations, Iran is divided into 8 regions and trip destination of Tehran residents is categorized in this regions. The ANOVA method is used for checking the effect of trip region, trip goal and trip time on the modal split. It has been shown that the 3 mentioned parameters can affect trip modal split. Analysis of gathered data shows that trip goals can be classified in the below categories: recreation, with 48%, meeting the relations and friends, with 19%, pilgrimage, with 18%, work, with 10%, and other goals including studying, business, and treatment with 5%. Using the generalized log linear models, dependency of trip region, trip goal, and trip time on the use of private car for intercity trips is studied. Pair combination of trip goal and trip time parameters, shows that most of the residents prefer to use private cars for traveling to Tehran region for recreation. In addition, most of the residents prefer to travel to the North region by their private cars for recreation.