This article presents a rule-based approach to making predictions about other people's psychological experiences. In this approach, individuals are assumed to know a set of hierarchically ordered transformation rules that they use to predict what psychological experiences (i.e., cognitions or affect) target persons will experience in reaction to stimuli. In Study 1, high school students made predictions regarding other people's cognitive and affective experiences in reaction to self-generated visual or ideational stimuli. A taxonomy of 10 transformation rules was derived to account for these predictions. The majority of rules was used for predicting both affective and cognitive experiences, but 4 were more frequently used for cognitive experiences and 2 were more frequently used for affective experiences. The source of the stimulus (i.e., visual vs. ideational) had little effect on rule use. Several analyses indicated that the observed pattern of rule use was consistent with a fixed-order hierarchy, and there was some evidence for individual differences in rule use. A replication study with university students found a very similar pattern of transformation rule use. A third study, in which high school students were asked to rate the likelihood of use of exemplars of each transformation rule in response to different stimuli, indicated that likelihood ratings were predictable on the basis of rule use in Studies 1 and 2. As well, hierarchical order in transformation rule use was demonstrated within subjects. The findings provide strong evidence regarding the theoretical and heuristic value of the transformation rule approach for predicting other people's psychological experiences.