Background Insufficient or excessive sleep duration are associated with increased risk of individual adverse outcomes. However, it remains largely unknown whether sleep duration trajectories are associated with overall health among older adults. This study aimed to examine the association between sleep duration trajectories and successful aging. Methods In the China Health and Retirement Longitudinal Study (CHARLS), 3,306 participants without major chronic diseases at baseline and survived to aged 60 years and older at the end of follow-up were potentially eligible participants. Total sleep duration was assessed in 2011, 2013, and 2015, and successful aging was evaluated in 2020 and was defined as being free of major chronic diseases, no physical impairment, high cognitive function, good mental health, and active engagement with life. Latent class mixed model (LCMM) was used to identify sleep duration trajectories and logistic regression was performed to explore the association between these trajectories and successful aging. Results During the 9-year follow-up, 455 individuals (13.8%) met the criteria for successful aging. Five sleep duration trajectories were identified: normal stable, long stable, decreasing, increasing, and short stable. Compared with the normal stable trajectory, the adjusted ORs (95% CI) for achieving successful aging for participants with long stable, decreasing, increasing, and short stable trajectories were 1.00 (0.77, 1.30), 0.64 (0.40, 1.03), 0.64 (0.45, 0.92), and 0.48 (0.35, 0.66), respectively. The stratified and sensitivity analyses were generally consistent with the main results. Conclusions Increasing and short stable trajectories of sleep duration are associated with lower odds of successful aging relative to participants in the normal stable trajectory. The findings underscore the critical importance of monitoring dynamic changes in sleep duration in middle-aged and older Chinese adults.