Four topics were investigated in this longitudinal person-centered study: (a) profiles of subjective task values and ability self-concepts of adolescents in the domain of mathematics, (b) the stability of and changes to the profiles of motivational beliefs from Grade 7 to 12, (c) the relation of changes to student-perceived classroom characteristics, and (d) the extent to which profile membership in early adolescence predicted mathematics achievement and career plans in late adolescence and the choice of math-related college majors and occupations in adulthood. Data were drawn from the Michigan Study of Adolescent and Adult Life Transitions Study. We focused on students who participated in the following 4 waves of data collection (N = 867): at the beginning of Grade 7 (Wave 3), at the end of Grade 7, in Grade 10 (Wave 5), and in Grade 12 (Wave 6). Four profiles that were stable across Grades 7 to 12 were identified using Latent Profile Analysis. Student-reported fairness and friendliness and competition in class predicted changes in profile membership. Profile membership in Grade 7 predicted math-related career plans in Grade 12. Profile membership in Grade 12 predicted the choice of math-related college major after finishing school and of math-related occupations in adulthood. Educational Impact and Implications Statement Findings of this longitudinal study with 867 students followed from the beginning of Grade 7 to adulthood showed 4 stable motivational subtypes: students with (a) high, (b) medium, (c) low motivational beliefs, and (d) students with moderate math self-concept and importance value, but low interest in math. Negative motivational changes in early adolescence were buffered if students perceived their teachers as friendly and fair. The motivational subtypes predicted the choice of math-related college majors and occupations in adulthood. By identifying different motivational subtypes among adolescents, findings emphasize that classrooms are characterized by high motivational heterogeneity of students that needs to be addressed in instructional settings. If these findings can be replicated in other studies, teachers should consider implementing personalized tasks that match in the motivational orientation of individual students.