The primary objective of this paper is to estimate the influence of Lake Erie water quality on the housing price by taking spatial effects into account. The robust LM tests for spatial autocorrelation suggested that spatial error model specification is more likely model in our study. Fecal coliform counts and Secchi depth disk reading are used as water quality measures. In order to overcome the spatio-temporal aspects of Secchi depth disk reading data, Kriging was used for spatial prediction. We found the significant influences of both water quality measures on housing values. Gradient effects considering the distance from a beach and water quality variables are also observed.