This paper explores the impact on beach attendance of beach closures and the intersite and intertemporal substitution that may follow beach closures. A model of beach attendance is developed that builds on a model constructed by Paul Ruud to support the State of California's claim to damages after the American Trader oil spill off the coast of Orange County, southern California. Newly gathered data on beach closures is combined with data on daily attendance from 1985-1993. Variables are constructed to test for intersite substitution (the shifting of beach recreation in space, i.e. from a closed beach to another beach) and intertemporal substitution (the shifting of demand for recreation at a particular beach over time). The method of non-linear least squares is used to estimate a system of five seemingly unrelated regression equations. For each equation, Breush-Pagan tests for heteroskedasticity fail to reject the null hypothesis of homoscedasticity and modified Breush-Godfrey tests for autocorrelation fail to reject the null hypothesis of no autocorrelation. The analysis produces only weak evidence to support rejection of null hypotheses that there are no effects due to beach closures, intertemporal substitution, or intersite substitution. For example, just two of six coefficients on closure variables are statistically significant. The lack of stronger evidence of casual effects likely reflects at least in part the fact that (1) the attendance data that form the foundation for analysis only extend from December to March and (2) people can still visit the beach when it is "closed" since the closures considered here pertain only to water contact. Such closures will likely have a greater effect during summer when air and water temperatures are higher and more people will want to engage in water-based recreation.