Stressful experiences are generally thought to be harmful to mental and physical health. This perspective has guided research, informed policy, and colored the public narrative on the effects of caregiving for an impaired older adult. Indeed, many studies find greater depression, higher negative affect, and lower life satisfaction in caregivers (CGs) relative to non-caregivers (non-CGs).1,2 However, there is evidence that CGs experience positive effects despite, or perhaps because of, their stressful experiences.3 Research also finds that CGs live longer and are more resilient to perceived stress than non-CGs.4,5 Moreover, many CGs report psychological benefits like increased meaning in life, growing closer to the care recipient, and learning new skills.6,7 In a previous paper,8 the PI (Marino) suggested that these benefits reflect domains of eudaimonic well-being (EWB - meaning in life, positive relationships, personal growth, environmental mastery, autonomy, and self- acceptance) developed in positive psychology research9. Longitudinal change in EWB among CGs of impaired older adults has not yet been studied. Higher EWB has also been linked to lower reactivity to stressors,10,11 but has not been evaluated as an internal resource that promotes CGs' resilience to daily stress. The proposed project will fill these gaps by examining links between caregiving status and multiple dimensions of well-being including EWB, hedonic well-being (HWB - positive and negative affect and life satisfaction), and affective and physiological reactivity to daily stress. We will obtain a sample of propensity-matched CGs and non-CGs from the main survey of the population-based Midlife in the United States study and its daily diary study (the National Study of Daily Experiences) to identify within- and between- group differences in 1) EWB and HWB, cross- sectionally and over 20 years, and 2) exposure and reactivity to stressors over 8 days. In an exploratory aim, we will apply novel machine learning techniques to identify diverse conditions associated with typologies of optimal and sub-optimal well-being. The central hypothesis is that despite being associated with persistently lower HWB, caregiving for an impaired older adult is linked to greater increases in EWB over time. Further, longitudinal gains in EWB are predicted to weaken the association between stressor exposure and reactivity to daily stressors in CGs. The project is expected to help shift the current CG well-being paradigm by applying robust methods from developmental psychology to the caregiving context and generating novel information about how caring for an aging adult might affect well-being both negatively and positively, concurrently and over time. Findings can inform new psychological interventions that target EWB as a pathway to increasing resilience to daily stressors. The proposed training site will provide a scientific environment that is challenging and strongly supportive of the proposed proj...