Annual income data are typically provided with a time lag. This article reviews several ways of dealing with this time lag in the construction of annual household-based income measures for individual economic well-being. It also proposes an alternative method that yields better estimates for equivalized household income, especially in the case of household composition change. Next, the two most commonly applied income measures are compared to this alternative measure with empirical income data from the European Community Household Panel. This comparison reveals that ignoring the time lag and household changes leads to substantial bias in income and poverty estimates and to erroneous conclusions about the determinants of poverty entry. The evidence in this article will be useful to researchers who want to make a well-informed choice between different annual income measures. © 2008 The Authors.