This research develops a theoretical model to analyze commuters' joint decisions of route and departure time in a simple network with two parallel routes: one with free but congested lanes, the other with free-flowing tolled lanes with time-varying tolls. By accounting for trip distance, the theoretical model is able to examine the two different sources of travel time uncertainties: length of commute and type of route. Length of commute causes long-distance commuters to leave home earlier, but type of route may lead to a change of route as well as of departure time to compensate for the higher risk of arriving late for work. Results show that long-distance commuters are less likely to take toll lanes than people with short commutes; that travelers who plan to take toll lanes are unlikely to switch to free lanes en route because their later departure times reflect the savings and reduced uncertainty in travel time; and that provision of pretrip information favors the use of toll lanes and may result in commuters leaving home later, even in heavier-than-normal traffic. The theoretical model provides more flexible demand specification for simulation studies. In addition it has important policy implications because it allows a better understanding of commuters' travel behaviors on roads with value-pricing schemes, such as the express lanes on CA-91 and high-occupancy toll lanes on I-15 in California.