Objectives Personal network turnover, a combination of lost and added network contacts, is suggested to affect health as well as moderate access to social support and resources. This article tests whether the caregiving process is associated with network turnover in later life and whether the process is different for men and women. Methods Network turnover was assessed using two waves of personal network data from the National Social Life, Health, and Aging Project. Network contacts were uniquely identified in each wave making it possible to document contact loss and addition. Rates of change were modeled using Poisson regression. Results Respondents transitioning into caregiving lost and added network contacts at higher rates than non-caregivers. Conversely, respondents providing care during both waves and respondents transitioning out of the role saw no significant levels of network turnover. The analysis provided minimal evidence of gender differences. Discussion Findings suggest that the initial shift into the caregiving role is associated with notable personal network change. This is an important consideration given that long-term network instability may lead to poor health and limited access to social resources whereas adaptive network change tends to elicit more positive outcomes.