No study has used trajectories of dietary patterns to examine their effects on sleep duration and body mass index over time in the Chinese population. We analyzed data from adults participating in the China Health and Nutrition Survey between 1991 and 2009. Dietary intake was measured by a 24-h recall method over three consecutive days. Height and body weight were measured, and sleep duration was self-reported. Multivariable mixed linear models were applied to examine the association between trajectories of dietary patterns (using a latent class model) and sleep duration as well as BMI. Four trajectories of a traditional pattern (characterized by rice, meat, and vegetables) and three trajectories of a modern pattern (characterized by fast food, milk, and deep-fried food) were identified. Participants with a high and rapid increase trajectory of the modern dietary pattern had the shortest sleep duration (β = -0.26; 95% CI: -0.40, -0.13). Participants with a high and stable intake of the traditional dietary pattern had the lowest BMI (β = -1.14; 95% CI: -1.41, -0.87), while the participants with a high and rapid increase trajectory of the modern dietary pattern had the highest BMI (β = 0.74; 95% CI: 0.34, 1,15). A rapid increase in the modern dietary pattern is associated with shorter sleep duration and higher BMI.