This paper examines the extent to which the conclusions of cross-country studies of private savings are robust to allowing for the possible heterogeneity of savings behaviour across countries and for the inclusion of dynamics. It reviews the econometric implications of neglected slope heterogeneity and dynamics for the fixed effects estimators routinely used in such studies, and illustrates the nature and extent of the biases involved by a re-examination of time series data from 21 OECD countries. The paper shows that neglecting heterogeneity and dynamics in cross-country savings regressions can lead to misleading inferences about the key determinants of savings behaviour. The results indicate that among the many variables considered in the literature only the fiscal variables - the general government surplus as a proportion of GDP and the ratio of government consumption to GDP - seem to be the key determinants of private savings rates in the industrial countries in the post-World War II period