This study demonstrates how cointegration analysis of privately-owned housing within disparate areas of the United States can aid developers in anticipating changes in the level of market activity. The study analyzes change in the number of housing units within four geographic regions: the Northeast, the Midwest, the South and the West. Whereas most studies of regional variation in real estate activity have focused on short-run analysis, this research extends the examination to consider the impact of exogenous variables over a longer time frame. The study uses Citibase data from 1959 through 1995. Results indicate that the four regions move together in the long run and are driven by one common factor, but that change in the South and the West lead those in the other two regions. Results have widespread policy implications for residential and commercial developers nationwide, because change within the dominant areas may serve as indicators of developing change elsewhere.