An inflow of immigrants into a region impacts house prices in three ways. For a fixed level of local population, housing demand rises due to the increase in foreign- born population. In addition, immigrants can influence native location decisions and induce additional shifts in demand. Finally, changes in housing supply conditions can in turn affect prices. Existing reduced form estimates of the effect of immigration on house prices capture the sum of all these effects. In this paper I propose a methodology to identify the different channels driving the total effect. I show that, conditional on supply, total changes in housing demand can be decomposed into the sum of direct immigrant demand and indirect demand changes from relocated population. The size and sign of the indirect demand effect depends on the impact of immigration on native mobility. I use Spanish data during the period 2001-2012 to estimate the different elements of the decomposition, applying an instrumental variables strategy to obtain consistent coefficients. The results show that overlooking the impact of immigration on native location induces a sizeable difference between the total and the immigrant demand effects, affecting the interpretation of the estimates.