With the aging of the "Baby Boomer" cohort, more and more adults are transiting from work into retirement. In public discourse, this development is framed as one of the major challenges of today's welfare societies. To develop social innovations that consider the everyday lives of older people requires a deeper theoretical understanding of the retiring process. In age studies, retiring has been approached from various theoretical perspectives, most prominently disengagement perspectives (retirement as the withdrawal from social roles and responsibilities) and rational choice perspectives (retiring as a rational decision based on incentives and penalties). Whereas, the former have been accused of promoting a deficient image of aging, the latter are criticized for concealing the socially stratified constraints older people experience. This paper proposes a practice-theoretical perspective on retiring, understanding it as a processual, practical accomplishment that involves various social practices, sites, and human, as well as non-human, actors. To exemplify this approach, I draw upon data from the project "Doing Retiring" that follows 30 older adults in Germany from 1 year before to 3 years after retirement. Results depict retiring as a complex process of change, assembled by social practices that are scattered across time, space, and carriers. Practice sequences and constellations differ significantly between older adults who retire expectedly and unexpectedly, for example through sudden job loss or illness. However, even among those who envisaged retiring "on their own terms," the agency to retire was distributed across the network of employers, retirement schemes, colleagues, laws, families, workplaces, bodies and health, and the future retiree themselves. Results identified a distinct set of sequentially organized practices that were temporally and spatially configured. Many study participants expressed an idea about a "right time to retire" embedded in the imagination of a chrononormative life-course, and they often experienced spatio-temporal withdrawal from the workplace (e.g., reduction of working hours) which entailed affective disengagement from work as well. In conclusion, a practice-theoretical perspective supports social innovations that target more than just the retiring individual.