Omission of substitute sites in travel cost analysis can cause an overestimation of recreational benefits. Only few analyses have included substitutes, partly because of the difficulty in defining an appropriate set of substitutes. We examine factors affecting the existence of substitutes and their impact on the demand and value of coastal recreation using spatially-referenced survey data from Finland, Germany and Latvia on recreational visits to the Baltic Sea, which can be characterized as a unique destination. Substitutes are defined by respondents themselves. Our findings indicate that the existence and effects of substitute sites differ across countries. Many respondents have no substitutes for Baltic Sea recreation, in particular in Latvia. Respondent and visit-specific factors explain the probability of having substitutes. Substitutes reduce the demand for coastal recreation in Finland and Germany but increase it in Latvia. Further, respondents having substitutes are less sensitive to travel costs in Germany and more sensitive in Finland and Latvia. The annual welfare from Baltic Sea recreation is lower for people who have substitutes in Finland and Germany, and higher in Latvia. The findings suggest that uniform assumptions about the existence and effects of substitutes appear unwarranted, especially for sites of unique or iconic nature.