A Discrete Choice Travel Cost model, based on data collected from a survey of recreational anglers, was used to estimate changes in recreational fishing benefits in the Upper Oldman River region of Alberta resulting from the construction of a dam. The results show that this model is useful for measuring the impact of public works projects on non-market economic benefits. Predictions of the distribution of trips to each fishing site before and after placement of the dam are also identified by the model. The model is sensitive to the variables chosen and the measurement of quality attributes. The quality attributes which affect the choice of site include the potential to catch fish (catch rate and size of fish), access, and the size of the water body. Including the value of travel time in the travel costs causes an increase in the welfare estimates, but does not affect the variables used. Construction of the dam and creation of the reservoir reduces recreational fishing benefits of the area. The welfare impacts of the decline of recreational fishing quality range from an annual loss of $96,239.10 to a loss of $30,545.20 depending on the model specification, and whether the value of time is included. The government efforts at mitigating the dam's effect by 'construction of fish habitat in remaining reaches may improve the welfare of users to levels equal to or greater than the original benefits. The mitigation effort, assuming a success rate that is considered most probable, results in an annual gain in welfare of from $209,499.80 to $22,971.60 depending on the model specification, and whether the value of time is included.