Recently, the importance of ill-structured problem-solving in real-world contexts has become a focus of educational research. Particularly, the hypothesis-development process has been examined as one of the keys to developing a high-quality solution in a problem context. The authors of this study examined predictive relations between young adolescents' metacognition, prior domain knowledge, and hypothesis-development performance in a computer-supported environment. Data were collected from 11- and 12-year-old Korean students (N = 101). A hypothesised model in predicting hypothesis-development performance was evaluated using structural equation modelling. Both metacognition and prior domain knowledge significantly predicted young adolescents' hypothesis-development performance. Implications and limitations of the present study and issues, including the experimental design, are discussed. (Contains 3 figures and 1 table.)