Using a sample of 678 property portfolio changes (acquisitions, dispositions and joint ventures) of U.S. REITs during the period 1990 to 2009, I investigate the issue of what drives the property sector focus of REITs. Geltner and Miller (2001) argue that investors prefer to make their own diversification decisions using narrowly focused REITs as an explanation for the lack of diversification. On the basis of their argument, I develop and examine the research question of how investors react to a change in a REIT’s property type focus. I find a significantly negative market reaction to acquisition and acquisitional JV events that decrease property-type focus. However, I do not find consistent supporting evidence that dispositional events, including property sales and dispositional JVs which increase property-type focus, yield significantly positive abnormal returns. Only in the limited case of other property-type dispositional JVs do I find a statistically significant positive market reaction relative to those derived from the dispositional events that do not change the property-type focus on the basis of a difference test. In terms of the results of cross-sectional OLS regressions, I also find strong evidence of a diversification discount derived from acquisitional events that decrease the property-type focus of a REIT regardless of the sample period and the type of property portfolio change. However, I do not find evidence of a wealth benefit received by dispositional events which increase the property-type focus. In addition, I find that the deal size of the property portfolio change relative to the size of the firm and the number of security analysts following the firm are both significant variables that affect the abnormal returns upon the announcement of a property portfolio change. I also find no evidence to support the idea that the diversification discount comes from endogeneity as argued by Villanova (2004).