The purpose of this study was to examine effects of model-based teaching on students' argumentation skills. Experimental design guided to the research. The participants of the study were pre-service physics teachers. The argumentative intervention lasted seven weeks. Data for this research were collected via video recordings and written arguments. Results show that construction of concrete models and using them in their discussions and explanations provide learners with more quality (accurate, consistent, appropriate, and relevant) arguments. In addition, models' quality affects the number of claims, evidences and reasoning that are produced during argumentation. The closer learners' models are to the real situations, the more argument components they generate.