Rhesus macaques (Macaca mulatta) are gregarious primates that form despotic societies characterized by frequent and intense aggression. Within long-term social groups, demographic changes may influence hierarchical stability, potentially resulting in conflict and violently abrupt hierarchical changes. This conflict can result in serious implications for animal welfare, and thus, predictive tools would be invaluable to captive managers in determining social instabilities. Using the method Elo-rating to track rank changes and dominance stability, we predicted that demographic changes to a population of semi-free ranging rhesus macaques would result in changes in hierarchical stability. Over a 3 y period, dominance data were recorded on all troop members to track the hierarchy. Throughout the 3 y, significant changes occurred to the population (mainly due to health and colony management reasons; no changes specifically occurred for this study) including permanent removal of a large group of natal males, temporary and permanent removal of top-ranking females, and depositions of top-ranking families. Our retrospective study suggests that removing natal males was beneficial in promoting overall troop stability (that is, stability of dominance relationships), although remaining males opportunistically attempted to increase in rank, perhaps due to limited competition. Our results also suggest that removing top-ranking females, even temporarily, destabilized dominance relationships; consequently adjacently ranked females opportunistically increased in Elo-rating, both before and after the depositions of the α families. Thus, these challenges to the established hierarchy can be predicted by increases in Elo-rating within the β families after demographic changes to the α families. Our results suggest that the presence of natal males and the removal of top-ranking females should be minimized to maintain stable dominance relationships. In addition, longitudinal data reflecting dominance ranks, collected by using Elo-rating, may help managers of captive colonies in predicting dominance instabilities before they occur.