Purpose: This study aimed to identify latent classifications based on trajectories of change in depressive symptoms among multicultural adolescents in middle and high school and to identify factors that influence classification. Methods: This study used six time-point data from the Multicultural Adolescents Panel Study (MAPS) conducted by the National Youth Policy Institute (NYPI). We longitudinally analyzed data from 2014 to 2019, which corresponds to the participants' middle and high school years. Latent Growth Curve Modeling (LGCM) was used to analyze the depressive developmental trajectories of multicultural youth in middle and high school. Latent Class Growth Modeling (LCGM) was used to determine the number and shape of latent classes according to the type of depression among multicultural adolescents. We applied multinomial logistics regression analysis to explore the predictors of each class. Results: A model with three latent classes was identified: (1) low-decreasing class (intercept=13.34, slope=-0.07, 37.9%), (2) moderate-increasing class (intercept=17.76, slope=0.41, 47.6%), (3) high-increasing class (intercept=23.22, slope=0.44, 14.5%). Multinomial logistic regression showed multicultural adolescents who were male, experienced high acculturative stress and social withdrawal, and had low self-esteem, life satisfaction, and family support tended to belong to either the moderate-increasing or high-increasing class. Conclusion: These findings suggest that preliminary assessments of gender, acculturation stress, self-esteem, life satisfaction, social withdrawal, and family support among multicultural adolescents may be valuable for developing tailored mental health programs.