Understanding one another’s preferences, e.g., for something or someone over others, is important for us to navigate the social world. Even infants seem to make sense of agents’ actions in terms of preferences, which are defined as dispositional states that help explain why an agent chooses a particular object in the presence of another option. So far, results on infants’ understanding of preferences have come from experimental situations in which the available evidence is straightforward, that is, when an agent’s choice between two options is consistent. I will present data showing when the agent chooses inconsistently between two options, infants can use different consistency/inconsistency ratios to learn about the agent's preferences. In addition, I will present data showing that infants can also engage in transitive reasoning about preferences. Together, these results demonstrate how general learning mechanisms, statistical learning and transitive reasoning in particular, inform infants' understanding about agents' preferences.