When people compare with another person they can focus on how they compare either with just that target (a personalized comparison) or with others more generally (a generalized comparison). Four studies (two event-contingent diary studies, one study of comparisons during a triathlon, and one controlled experiment) showed that personalized comparisons were more likely when the target's attribute was distinctive or there was an interaction or a close or emotional relationship with the target. Perhaps because these conditions that increase interest in the target as a distinct individual were less common during the triathlon than in everyday life, personalized comparisons were relatively uncommon during the triathlon but relatively common in everyday life. Across studies, generalized comparisons magnified the impact of upward comparisons on overall feelings (but not on interpersonal feelings about the self-target relationship), presumably because generalizing broadens the implications of comparisons, whereas personalizing restricts their relevance to the self-target relationship.