The article develops models to estimate the benefits of recreational fishing by combining real and hypothetical data. By using both types of data, the researchers found more accurate estimates of the effects of trip costs and characteristics on anglers' choices. The joint estimation of revealed and stated preference data led to significant improvements in efficiency and reduced differences in willingness to pay estimates. Different models were compared, with the mixed logit error components model being preferred for its ability to account for panel data and scale differences. Overall, the study shows that combining different types of data can provide a more comprehensive understanding of recreational benefits.