Social support is typically thought to protect people from bad events, whereas capitalization support augments people’s reactions to good events. Because social support and capitalization support apply to different classes of events, most theory predicts that measures of perceived support and capitalization support should be empirically distinct. We tested a new theory that hypothesizes that the main effects between perceived support and mental health do not reflect stress and coping primarily, but instead reflect ordinary, yet affectively consequential conversations and shared activities, some of which include positive events. According to this view, perceived support and capitalization support should be substantially correlated, should have similar links to other constructs, and their links to favorable affect should overlap, yet not be completely redundant. In three samples, results were consistent with the new theory, when correlations reflected social influences. When correlations reflected trait influences, perceived and capitalization support showed greater overlap.