In an inquiry-learning task, computer modeling can be a powerful tool to enhance students’ reasoning and help them improve their understanding. Optimal learning effects in such an environment require a high quality of students reasoning activities. It is not straightforward, however, which type of modeling tool will be most helpful to students’ reasoning. In order to identify the effects of different tools, students working with two common types of tools were compared to a normative description of the modeling process. Also the influence of reasoning activities on the achieved modeling result was examined. Different modeling tools did induce significantly different reasoning activities. Students working with a graphical representation designed more experiments with their own model, formulated more qualitative hypotheses, and spent more time evaluating their own model than students working with a textual representation. Results also indicate that many students have serious difficulties performing this task in a systematic manner. The paper concludes with suggestions for support students might need.