This paper considers two contenders for the title of highest good in Kant's theory of practical reason: happiness proportioned to virtue and the maximization of happiness and virtue. I defend the ‘proportionality thesis’ against criticisms made by Andrews Reath and others, and show how it resolves a dualism between prudential and moral practical reasoning. By distinguishing between the highest good as a principle of evaluation and an object of agency, I conclude that the maximization of happiness and virtue is a corollary of the instantiation of the proportionality thesis.