The aim of this paper is to test whether Google search volume indices can be used to predict house prices and to identify bubbles in the housing market. We analyse the 06/07 U.S. housing bubble, taking advantage of the hetrogenius house price development in different U.S. states with both bubble and non-bubble states. From 204 housing related keywords, we test both single search terms and indexes with sets of search terms and finds that the several keywords preforms very well as a bubble indicator. Google search for Real Estate Agent displayed the most predictive power for the house prices, of all the keywords and indexes tested, globally in the US. Google searches volume outperforms the well-established Consumer Confidence Index as a leading indicator for the housing market.