A one-year longitudinal study was conducted to investigate the accuracy of people’s assessmentsof their own personal change over time. We compared people’s predicted, actual, and recalledchange in their personality, values, and preferences over this time period. On average,participants underestimated the absolute magnitude of their personal change, yet simultaneouslyoverestimated their net improvement, in both prediction and recall. This effect was due to anasymmetry whereby people selectively neglected negative changes, especially prospectively.Although participants in our sample both improved and declined over the year, they were morelikely to remember past improvements than declines, and made nearly uniformly positivepredictions of future change. We discuss how the current findings reconcile researchdemonstrating expectations of personal improvement (e.g., Wilson & Ross, 2001; Kanten &Teigen, 2008) with other research that suggests people overpredict their personal stability(Quoidbach, Gilbert, & Wilson, 2013).