Two theoretical perspectives on the connection between religion and politics are applied to Great Britain. Data from the 1991 and 1992 waves of the British Household Panel Study, used to conduct multinomial logistic regression analysis, dispute the general consensus that religion has weak or no effects on the voting decisions of British citizens. Religious belonging, behaviour and belief, as well as the religious context of households, continue to influence British voting behaviour. Interaction effects among religious variables and between religious variables and class also operate to influence vote choice. Areas for further research into the religious bases of British electoral behaviour are suggested.