Modeling is a core practice in science and a central part of scientific literacy. We present theoretical and empirical motivation for a learning progression for scientific modeling that aims to make the practice accessible and meaningful for learners. We define scientific modeling as including the elements of the practice (constructing, using, evaluating, and revising scientific models) and the metaknowledge that guides and motivates the practice (e.g., understanding the nature and purpose of models). Our learning progression for scientific modeling includes two dimensions that combine metaknowledge and elements of practice-scientific models as tools for predicting and explaining, and models change as understanding improves. We describe levels of progress along these two dimensions of our progression and illustrate them with classroom examples from 5th and 6th graders engaged in modeling. Our illustrations indicate that both groups of learners productively engaged in constructing and revising increasingly accurate models that included powerful explanatory mechanisms, and applied these models to make predictions for closely related phenomena. Furthermore, we show how students engaged in modeling practices move along levels of this progression. In particular, students moved from illustrative to explanatory models, and developed increasingly sophisticated views of the explanatory nature of models, shifting from models as correct or incorrect to models as encompassing explanations for multiple aspects of a target phenomenon. They also developed more nuanced reasons to revise models. Finally, we present challenges for learners in modeling practices-such as understanding how constructing a model can aid their own sensemaking, and seeing model building as a way to generate new knowledge rather than represent what they have already learned. ß 2009 Wiley Periodicals, Inc. J Res Sci Teach 46: 632-654, 2009 Keywords: scientific modeling; learning progression; scientific practice; student learningResearch-based reforms in science education have emphasized the importance of engaging learners in scientific practices-social interactions, tools, and language that represent the disciplinary norms for how scientific knowledge is constructed, evaluated, and communicated (Duschl, Schweingruber, & Shouse, 2007). Involving learners in developing and investigating explanations and models leads to more sophisticated understanding of key models in science, as well as helping learners understand the nature of disciplinary knowledge (e.g., Lehrer & Schauble, 2006). Yet, scientific practices require shifts in traditional classroom norms that involve learners in knowledge building and negotiation (Berland & Reiser, 2009; Jimenenez-Aleixandre, Rodriguez, & Duschl, 2000;Lemke, 1990). For effective participation in scientific practices, teachers and students need support with the practices as well as with the scientific ideas addressed by the practice (Duschl et al., 2007).The MoDeLS project, Modeling Designs for Learning Scien...