This study examines the links between formal community participation (defined as volunteer work for an organization) in the United States and psychological well-being, focusing on retirement as a key status transition. It draws on data from 762 American retirees and not-yet-retired older workers in the midcourse years (in their 50s, 60s, and early 70s) to examine: (1) whether unpaid community participation of retirees compensates for the loss of connectedness associated with their paid career jobs, and (2) whether community participation predicts the well-being of various subgroups of retirees. We find little difference in the community participation of midcourse Americans by their retirement status. But community participation does predict the well-being of retirees, supporting a compensatory hypothesis. Dividing the sample of retirees into those who engage in post-retirement paid work and those who do not reveals a positive association between community participation and well-being for non-workers only. Community participation appears especially salutary for the well-being of some subgroups of retirees in the United States (men more than women, those with less income, and those in poor health).