Abstract Using home sales data for 10 years and four sites in Maine (USA), we estimate marginal willingness to pay (WTP) for river proximity and present a criterion-validity test of the use of benefit transfer. Given the degraded state of Maine's waterways (partly due to damming of rivers for commerce), we find positive, concave relationships between housing values and river distance. We find statistically different estimates of marginal WTP across the four markets, but results from the same watershed are more similar. Also, we find mean WTP is diminishing with average river distance, which is consistent with the empirical functional form. Comparing original and benefit transfer estimates, we find errors range from 29% to over 1000%, growing with spatial distance between the study and transfer sites. Lastly, we estimate a difference-in-difference model for the single market that has witnessed dam removal, and we apply the results to the other markets (that could undergo dam removal as well). While informative, our results suggest caution in the use of benefit transfer in applications of hedonic property price analysis.