Abstract Computational experiment approach considers modelling as the essential feature of Inquiry Based Science Education (IBSE), where the model and the computer take the place of the “classical” experimental set-up and simulation replaces the experiment (Landau, Pαez, & Bordeianu, 2008). Modelling, as a pedagogical tool, involves the model construction, the exploration of model characteristics and the model application to a specific problem, resembling authentic activities of scientists and mathematicians (Herbert, 2003). Recent developments in strategy instruction research suggest that learning in a particular discipline is enhanced by guiding students through the development of content-relevant metacognitive strategies (Wosnitza & Volet, 2009). Problem-solving is a complex process, which involves several cognitive operations such as collecting and selecting information, heuristic strategy and metacognition (De Corte, 2003; Garofalo & Lester, 1985; Schoenfeld, 1994). The purpose of this study was to explore the impact of the Computational Experiment Methodology on learners' cognitive performance, use of modelling indicators and shift of the metacognitive experiences during problem solving using computational models. Sixty prospective primary school teachers volunteered to participate in the study. Students were exposed by the Instructor to a number of computational experiments, while during the course they developed their own models of simulation. The results of the experiment show that the use of the computational experiment approach has a substantial effect on the metacognitive experiences and the use of modelling indicators.