In this study we examined change in students' situational interest as a function of student and task characteristics. Fifth- and sixth-graders (n = 52) were assigned to one of two task conditions that used a different version of a science simulation. The versions differed in how concrete vs. abstract the simulation elements were. Students' prior knowledge, achievement goal orientations, and subject-specific interest were assessed before the task and situational interest was measured repeatedly in different phases of the task. Post-task performance was assessed 1 day after the task. The results showed different mean-level changes in situational interest in the two task conditions; students working with the more concrete version of the simulation reported increase in their interest while the opposite was true for students working with the more abstract version. The ratings of situational interest were nevertheless rather stable over time, regardless of the task condition. Students' situational interest at the beginning of the task was predicted by mastery-intrinsic goal orientation and subject-specific interest. Post-task performance was predicted by prior knowledge and the task condition; students working in the more concrete task condition performed better. The importance of acknowledging both individual characteristics and task elements in the emergence of students' situational interest is discussed.