Approximately 40 percent of fuel consumption in large cities is related to transportation. A noticeable amount of fuel is wasted due to traffic congestion in peak hours.Transportation planners look for policies to reduce congestion to save fuel and increase energy efficiency. One of the policies is carpooling that emphasizes on a shared use of private cars. In this paper, the factors which persuade travellers to choose carpooling are investigated for Tehran city, capital of Iran. A stated preferences (SP) survey has been used to observe travellers’ tendency of carpooling. SP is a survey technique which mathematically shows the preferences, based on people's stated choices and their responses to hypothetical situations. The survey questionnaires filled out by 470 travellers used their own automobiles. Considering the data, carpooling impacts are analyzed in different situations. In this approach a demand function is calibrated and utilized to predict percent of travellers choose carpooling. When all interested travellers, independence of knowing appropriate rideshare or not, choose carpooling then vehicle trips per day would decrease about 780000 vehicle trips per day and reduce annual fuel consumption by 336.53 million litres. The results show that if appropriate strategies like carpooling websites are designed to help travellers for identifying appropriate rideshares, carpooling would increase by 30 percent and this increase will reduce annual fuel consumption about 240 million litres. Results also show that high occupancy vehicle lanes (HOV) that reduce travel time for ridesharing may not highly influence on carpooling tendency of travellers.