Malleable, noncognitive psychological factors such as grit and resilience are seen as critical for academic success and have garnered significant attention from researchers and policymakers. However, there is nontrivial overlap between these two constructs, and it remains unclear whether grit, resilience, or both, constitute the optimal target(s) for retention research and intervention development. The purpose of this study was to compare how grit and resilience predict retention in a diverse, multicollege sample of U.S. undergraduates (N = 4,023) with and without a history of academic probation. When looking across all students, structural equation modeling indicated that perseverance of effort (Grit-PE) positively predicted Fall 2016 enrollment. However, when examining probation and nonprobation groups separately, only resilience significantly predicted enrollment for students with probation histories. In contrast, Grit-PE predicted enrollment only for students without probation histories, while consistency of interest (Grit-CI) did not predict enrollment for either probation group. Findings held when controlling for race, gender identity, years in college, first-generation status, and institutional selectivity. Results suggest that resilience, not grit, predicts retention for those students who face the greatest setbacks in completing college.