ABSTRACTButler and Stokes’Political Change in Britainis reviewed, and criticised for focussing on a notion of electoral change which is singularly restrictive and which, moreover, has been noticeably absent in post‐war Britain compared with other Western democracies. Three other kinds of electoral change, although ignored in the book, are shown to have been peculiar to Britain since 1945. These are (i) a persistent decline in the combined major party share of the electorate; (ii) a gradual fall in turnout; and (iii) an accelerating volatility of support between the two main parties. A more disturbing weakness in the study is that its general model of electoral behaviour, largely borrowed from the earlier Michigan studies, cannot account for these three electoral trends. Indeed, the model leads to predictions of electoral behaviour which are the very opposite of what has in fact taken place. In the light of the model's failure, some of its key concepts such as party identification, political generation and the model of partisan development as a learning process are subjected to critical scrutiny. No alternative model is tested, but plausible explanations of British voting behaviour since 1945 are offered which place a much greater emphasis on sociological and historical factors and on changes at the macro and elite le