This study investigates whether the Feldstein and Horioka [Econ. J. 90 (1980) 314.] argument on domestic saving-investment relationship may remain as a "puzzle" when an endogenous structural break corresponding to a major policy regime change is taken into account. To this end, we employ not only the conventional procedures developed for a data generation process without a break, but also some recent methods which allow stationarity around an endogenous break under the alternative hypothesis. The evidence based on the UK annual data from 1948 to 1998 suggests that the long-run relationship between saving and investment disappears after the estimated endogenous structural break point coinciding with the abolition of foreign exchange controls in 1979. © 2002 Society for Policy Modeling. Published by Elsevier Science Inc. All rights reserved.