How should a group of agents make a decision when they face uncertainty about the consequences of possible actions? I study this problem when individual and collective preferences are based on subjective expected utility. A natural way to determine the preferences of a group then is to average its members’ beliefs and add up their (0,1)-normalized utility functions. This procedure extends the idea of relative utilitarianism to decision making under uncertainty. I show that it is the only way to aggregate preferences that responds monotonically to changes in the agents’ preferences if differences in preferences are due solely to differences in the utility functions and that satisfies a weak version of Arrow’s independence of irrelevant alternatives as well as four undiscriminating axioms.