Social support has been shown to enhance adjustment to retirement. Continuity theorists propose that retirees rely on internal dispositions and skills as a "continuity strategy" to maintain consistency in support and well-being despite disruptions to social networks, yet such skills have not been examined to date. Also, despite research indicating a disproportionate impact of negative support on well-being, retirement studies have relied heavily on measures of positive support. The present research employed structural equation modeling to assess "emotional awareness" as a putative socio-emotional skill underlying positive and negative support and well-being longitudinally, in recent retirees. The mediating role of perceived satisfaction with support was also examined. Two studies were employed to test three main hypotheses: (1) emotional awareness is a socio-emotional skill employed by retirees to maintain consistency in the frequency of positive and negative interactions; (2) negative interactions will have a stronger impact on well-being in retirement than positive interactions (i.e., a "negativity effect"), and (3) the impact of positive and negative interactions will be partially mediated by perceived satisfaction with support. In study 1, cross-sectional results indicated that emotional awareness facilitated positive, but did not mitigate negative interactions. Negative interactions had more potent effects on well-being than both positive interactions and perceived satisfaction with support, suggesting a strong "negativity effect". Also perceived satisfaction partially mediated the impact of positive and negative interactions on well-being. Study 2 assessed the aforementioned relationships longitudinally. Emotional awareness facilitated positive interactions in early retirement, leading to improved satisfaction with support and well-being longitudinally. Negative interactions exerted more potent effects on well-being than positive support, suggesting a strong "negativity effect" in retirement longitudinally. Perceived satisfaction with support did not mediate the relationships between positive and negative support and well-being longitudinally. Implications for research and intervention are discussed.