There is a strong current in the social sciences at the present time arguing that the concept of a separate urban phenomenon, defined spatially, is irrelevant to understanding society. Against this, others argue that places matter in a whole variety of ways. Recent analyses of voting behavior in Britain support the latter view, but they are based almost entirely on aggregate data. Analyses of a sample of 3,955 British voters, taken in 1983, provide further support for the latter view. Labour voting is shown to vary not only by class but also by place—both region and settlement type. The reasons for this are explored.