ABSTRACT Supporting students in complex problem-solving tasks is intricate due to the variety of difficulties students face. Although adaptive support has proven effective in addressing various difficulties in different disciplines, its application and exploration in science education, particularly through intelligent tutoring systems (ITSs), is limited. This study aimed to investigate the effectiveness of adaptive support in fostering students’ problem-solving processes in organic chemistry, focusing on deriving insights for designing an ITS. Adaptive support, in the form of adaptive stepped supporting tools (ASSTs), was developed and implemented in problem-centred interviews with 19 undergraduate chemistry students. The analysis indicated that ASSTs generally supported students’ problem-solving processes effectively. Additionally, certain factors were identified that influenced the effective processing of ASSTs. The results suggest that an effective ITS should feature highly flexible support pathways, a prior knowledge assessment, the explicit communication of identified difficulties, a drop-out option, and discipline-specific task formats. However, the results also raise questions about the assumed extent of prior knowledge in adaptive support and the point at which the amount of required support becomes impractical.