In order to provide meaningful traffic information to both travelers and traffic managers, it is critical to develop accurate and reliable traffic prediction algorithms that not only reduce absolute value of prediction error but also take into consideration the uncertainty associated with travel time prediction. The objective of this research is to identify and model uncertainties associated with travel time prediction and develop models for short term forecasting of the traffic state. Most existing travel time prediction methods only provide a point value as the prediction result which does not represent the uncertainty issues. Instead of providing a point value (an average of travel time during a certain time interval), a prediction interval based approach is proposed. The prediction interval represents likeliness of capturing true value of the future travel time. In other words, a prediction interval is an estimated range that captures the future observation, with a prescribed probability, given the current available observations.