<p>Different researchers prioritise different conversational processes as core to our relationships and our wellbeing. Two such processes are social reminiscing and emotional support. Social reminiscing and emotional support share conceptual and functional similarities, but previous research has not systematically compared them. Could they be the same? This thesis sought to disentangle social reminiscing from emotional support in several ways. In Chapter 2, participants imagined stressful scenarios (Study 1, N = 381) or recalled stressful and happy past events (Study 2, N = 380), and reported on the social reminiscing and emotional support that they desired (Study 1) or received (Study 2) from their conversation partner. Study 1’s results indicated that social reminiscing and emotional support are statistically distinct processes: People desired more reminiscing when they felt closer to the support provider, but desired consistently high levels of emotional support, regardless of how close they felt to the support provider. Study 2 produced more mixed results, finding no evidence that closeness changed how much reminiscing or emotional support was recalled in real life. Chapter 3 developed and validated a dyadic observational coding scheme for social reminiscing with a sample of people who discussed past events (a memory-directed context; Study 3, N = 51 dyads) and a sample of romantic partners who discussed future goals (i.e., a support-directed context; Study 4, N = 100 dyads). Results provided further evidence that social reminiscing and emotional support are distinct processes that work in tandem within a conversation. Finally, Chapter 4 included an observational study (N = 66 dyads) which examined the links between people’s depressive symptoms, dyadic patterns of social reminiscing and emotional support in conversations, and relational wellbeing. No evidence emerged linking depressive symptoms with patterns of social reminiscing and emotional support. However, discriminant findings indicated that social reminiscing predicted greater satisfaction with the conversation, but not perceived partner responsiveness. Conversely, emotional support predicted greater perceived partner responsiveness, but not satisfaction. Altogether, this thesis presents the first evidence that social reminiscing and emotional support can be statistically distinguished and are each a unique feature of conversations. Practically, this research suggests that in a real-world conversation, people respond qualitatively differently to past-tense and present-tense emotions. More inter-disciplinary work between relationship scientists and autobiographical memory researchers will be fruitful because their different theories are complementary.</p>