Evaluating the impact of tourism on housing prices is an important endeavour, but the usual empirical approach is to estimate a single regression model with house price as a function of tourism and other variables. This approach does not allow for individual heterogeneity. In this paper the authors apply a latent class model to estimate the impact of tourism activities on housing prices in Italy. In particular, they allow for three different unobservable classes or regimes, permitting the impact of tourism on house prices to differ across classes. In other words, they allow for unobservable heterogeneity. The empirical results do support the existence of three classes of house price regressions. Using two different indices of tourism activity, for certain cities (about 21% to 48% of the sample) increases in tourism activity increase housing prices and for other cities (8% to 17%) increases in tourism activity decreases housing prices. For approximately half of the sample, increases in tourism activity have no impact on housing prices.