A central concern in the field of international finance is always capital mobility. Feldstein and Horioka (1980) propose a simple test for international capital mobility and obtain a sign of very low capital mobility. Their interesting result is often described as the Feldstein-Horioka paradox. This paper reexamines their study using panel data analysis. Following the standard model selection procedure, preferred estimators of the elasticity of domestic investment-GDP ratio on domestic saving-GDP ratio are always significantly lower than one. In the light of our results, the Feldstein-Horioka paradox turns out to be not so robust because of cross country heterogeneities.