A simulation methodology for analyzing policies for reducing traffic congestion and the effect of providing commuters with pre-trip information is developed. This combines a model incorporating travel time uncertainty to determine optimal home departure times with a supply-side congestion model of a highway facility. Congestion is endogenously generated for different scenarios using a randomly generated normal distribution of work start times and previously estimated parameters. The results suggest that greater reductions in the expected costs of commuting may be achievable by policies that reduce travel time variance rather than just travel times. Policy simulations which analyze the congestion effects of providing commuters with pre-trip information are found to not necessarily provide many benefits to those using the information, although gains may be minor because the only behavioral response is limited to departure time changes.