Most previous research that models travelers' behavior in using managed lanes (MLs) versus a toll-free route has derived the individual's route-choice decision using a utility maximization approach. More recent models incorporating risk are based on expected utility theory (EUT). However, violations of some key assumptions of the EUT have led to the development of nonexpected utility theories, among which prospect theory (PT) has been one the most widely examined. This study examined if PT is superior to EUT when predicting route/mode choice and understanding travelers' behavior in the case of MLs by embedding PT proposed value function and probability weighting functions in the utility estimation. From both EUT and PT approaches, this study used survey data from 2012 to predict the mode choices that include MLs and toll-free alternatives, and provided estimates of the value that travelers are willing to pay (WTP) for travel time savings on MLs. The responses from the survey were examined using advanced discrete choice modeling techniques. Significant and interesting general findings resemble those in previous studies that use PT, including the fact that individuals weight probabilities. Two survey design methodologies, Db-efficient and adaptive random, were tested in this survey. Estimates from the EUT and PT approaches, as well as from previous studies on Katy Freeway travelers, are compared. The results of this study indicate that Katy Freeway travelers are more risk averse when in a situation of being late for work than they are with potential savings in travel time, and they, on average, demonstrate a sense of optimism when the chances of facing a longer travel time are high. PT based models, particularly the model embedding with probability weighting, outperforms EUT based models in terms of the predicative power. On average, models with probability weighting resulted in more than 65 percent of all mode choices correctly predicted, while conventional EUT models predict about 35 percent of choices correctly among four alternatives. Compared to previously available route choice studies, the relatively low willingness to pay (WTP) measures ($8 to $14/hour) calculated in this study from the PT models may deserve further investigation. Empirical findings from this study would help the policy makers set up appropriate project goals and toll rates to meet the increasing traffic demand of Katy Freeway travelers. The patronage of toll facility and MLs largely depends on the potential benefits (more reliable travel time and/or travel time savings) offered by such a facility. How the travelers actually perceive the potential benefits may have a significant influence on the use of MLs. This is about the belief that the travelers have on the facility. In lieu of the significant improvement in predicative power of the models embedding probability weighting functions and because of the stochastic nature of travel times, in future survey efforts it might be helpful to collect information regarding Katy Freeway travelers' actual belief on the benefits from using the MLs, and compare their 'belief' with the actual probability of reliable travel time and savings. Such comparison might help verify the accuracy of the probability weighting functions obtained in this study. The electronic version of this dissertation is accessible from http://hdl.handle.net/1969.1/152505