There are two kinds of beliefs. If the ultimate objective is wellbeing (utility), the generated beliefs are practical. If the ultimate objective is truth, the generated beliefs are scientific. This article defends the practical/scientific belief distinction. The proposed distinction has been ignored by standard rational choice theory?as well as by its two major critics, viz., the Tversky/Kahneman program and the Simon/Gigerenzer program. One ramification of the proposed distinction is clear: agents who make errors with regard to scientific beliefs (e.g., the conjunction fallacy) should not be taken as committing irrationality - because they are most probably engaging the other kind of maximization, the pursuit of wellbeing.