Recent efforts to identify non-cognitive predictors of academic achievement have especially focused on self-constructs, whose measurement is concerned with a specific domain (e.g., mathematics). However, other important factors, such as character and motivation, have received less attention. Additionally, the predictive accuracy of non-cognitive factors lacks evidence from subjects including English and Science. In this study, we take a person-centered approach and focus on students' intrinsic motivation, need for cognition, grit, and growth mindset. We mainly focus on how these factors predict students' mathematics, English, and science grades between 9th grade and 12th grade. 2,308 samples from high school students in Boston (Female = 1,237; aged from 13 to 17). The research results indicated that: (1) four latent profiles of students emerged: High in grit students (n = 997, 43.2%, higher scores of grit); Moderate students (n = 905, 38.3%, moderate in all scores); High in intrinsic motivation students (n = 252, 11.8%, higher scores of intrinsic motivation); Low in grit students (n = 154, 6.7%, lower scores of grit); (2) students' gender, race, maternal education level, and social-economic ranking predicted the profiles; and (3) four profiles of students had a significant predictive effect on Mathematics, Science and English scores in both 9th grade and 12th grade. We discussed the importance of character education for adolescents and motivation for learning in high school.