Problem-based learning (PBL), initiated when learners meet the problem scenarios, develops learners' skills and subject matter content to make the transition from beginner to more expert problem solver. Beginning students might benefit from simple, well-structured problems, in which the understanding of sets of rules and principles results in a straightforward solution. Facilitators might consider providing learners with more experience with complex, ill-structured problems. However, seldom have reports provided PBL facilitators the solutions for how to design or select problem scenarios so that they open the door to significant learning outcomes consistent with course or program goals. The gray area between well-structured problems and ill-structured problems needs to be further clarified for better learning effect. After a brief literature review, the author outlines a decision model for problem selection that addresses three dimensions: (1) conditions, including learning objectives, prerequisite skills and knowledge, available time, and presentation formats; (2) methods, including the simplifying conditions method and the problem-stimulated/student-centered strategy; and (3) outcomes for well-structured and ill-structured problems. The conclusion suggests that a PBL facilitator should consider the learning strategies both systematically and systemically. Future research is required to further develop this decision model. (Contains 29 references.) (MES)