Political parties can enjoy, as well as endure, remarkable reversals of fortune. The recent history of the British Labour Party provides an excellent example. In April 1992, Labour suffered its fourth consecutive general election defeat, and observers speculated that this might have been the party's "last chance" (e.g., Heath et al. 1994). Soon after that election, however, Labour support began to revive, and the governing Conservatives fell far and fast from public grace. Labour's resurgence was not ephemeral; rather, it continued and culminated in a spectacular victory in the May 1, 1997, general election.(1) In this paper, we analyze the dynamics of Labour's popularity surge by specifying new models of the political economy of party support. These models emphasize the short- and long-run effects of party leader images and party identification, as well as the influence of economic evaluations and emotional reactions to the economy, on Labour's vote intention share. The models thus depart from their predecessors, a large majority of which have relied heavily on a small set of macroeconomic indicators, especially inflation and unemployment rates, as principal explanatory factors (for reviews, see Lewis-Beck 1988; Miller 1989; Norpoth 1992). By incorporating variables that have received little or no attention in previous studies of aggregate movements in vote intentions, the new models are designed to help bridge the longstanding theoretical gap between such studies and individual-level analyses of electoral choice.