This study examines the use of question prompts and peer interactions as scaffolding strategies to help undergraduate students with their problem-solving processes on an ill-structured task. The mixed research method, combining both experimental and comparative multiple-case studies, was used to study the outcomes as well as the processes of students' problem-solving activities in terms of problem representation, developing solutions, constructing argumentation, and monitoring and evaluation. The result of the experimental study showed that the students working with peers and also receiving question prompts (PQ) significantly outperformed the other treatment groups, that is, individuals with question prompts (IQ), individuals without question prompts (IC), and peers without question prompts (PC). At the same time, though the students in the IQ group did less well than the PQ group in the process of problem representation, they significantly outperformed the PC and the IC groups in the processes of problem representation, justifications, and monitoring and evaluation. There were no significant differences between the PC and the IC groups in any of the four problem-solving processes. It appeared that question prompts were a superior scaffolding strategy over peer interactions in supporting students' ill-structured problem-solving processes. However, the comparative, multiple case studies revealed the complexity of the peer interaction context and the relationship between question prompts and peer interactions. While this study confirms previous findings on the effectiveness of question prompts in facilitating students' cognition and metacognition, it also indicates the benefits of peer interactions, which were contingent upon group members' active and productive engagement, that is, questioning, explaining, elaborating and providing feedback among peers. The study implies that, in order for students to gain full benefits from peer interactions, the peer interaction process itself needs to be scaffolded, especially when students were novice problem solvers; and question prompts, through expert modeling, may serve to facilitate this process. (Contains 23 references.) (Author/AEF)