The purposes of this article are to present the design of a technology-enhanced learning environment (Air Pollution Modeling Environment [APoME]) that was informed by a novice-expert analysis and to discuss high school students' development of modelling practices in the learning environment. APoME was designed to help high school students understand that air quality is the result of complex interactions among air pollutants, topographic effects, and meteorological variables. Using a novice-expert analysis, my research team specified atmospheric scientists' modelling practices as learning objectives, designed a dynamic modelling tool in light of the knowledge bases of scientists and students, provided students with dynamic simulations to help them visualise complex processes, and designed learning activities to encourage model-based reasoning. One teacher and 29 10th-graders participated in an implementation study that examined students' conceptual understandings and modelling practices in APoME. The results show that students' understandings about air quality were significantly improved after they engaged in the APoME activities. Students designed plans with detailed procedures, identified more major variables relevant to air pollutant dispersion, carefully controlled and manipulated variables to test their model, and provided multiple ways for data collection. These findings suggest that APoME is effective in supporting students to demonstrate expert-like modelling practices. (Contains 3 tables and 5 figures.)