This study investigated students' fixed mindset trajectories using 6-year longitudinal data from the 4th grade (elementary school) to the 9th grade (middle school). The longitudinal associations between the trajectories and academic achievement with behavioral engagement as a mediator were examined using a piecewise latent growth model. The Gyeonggi Education Panel Study with 3531 Korean students was used. Fixed mindset and school engagement were measured six times at 1-year intervals. The average score of the 9th-grade math, English, and Korean test scores were used as the outcome variable. The results showed that the fixed mindset slightly decreased during the elementary school period (4th-6th grade) but increased in the middle school (7th-9th grade). The initial level of fixed mindset negatively predicted academic achievement through the low initial level of behavior engagement (i.e., effort exerted, attention, and involvement in academics). The slope of a fixed mindset in the elementary school period did not influence academic achievement in the 9th grade. Conversely, the slope of a fixed mindset in the middle school period affected academic achievement through the slope of behavior engagement in the middle school period. The results indicated that trajectory patterns of fixed mindset were differentiated depending on the school level. An increase in students' fixed mindset in middle school plays a significant role in decreasing classroom engagement and academic achievement.