The purpose of this research project is to develop and examine a new handheld computer simulation platform designed to exploit the affordances of handheld technologies. We believe that a powerful handheld learning environment might capitalize on the portability, social interactivity, context sensitivity, connectivity, and individuality of ubiquitous devices to bridge real and virtual worlds. Our platform will enable the development of “augmented reality” simulations, that is simulations that provide a virtual context layered on top of a real-world context. The handheld computer then provides a window into the virtual context that is sensitive to information being supplied to it by the real world. Our first step was to create a proof of concept module representing the types of simulations that the development environment will create. The prototype, “Environmental Detectives,” is a participatory simulation where groups of students participate in a real-time simulation-game based around a local watershed. This real- world situation is augmented by a simulation of an environmental disaster; in this case a toxic spill that can potentially contaminate ground and surface water. The handheld provides a window into that simulation where students can take sample readings, interview people and access geographical information. To solve the problem, students must combine real-world and virtual-world data and develop an argument about how to best remediate the problem. The spread of TCE is simulated on a location-aware Pocket PC equipped with a GPS device, which allows players to sample chemical concentrations in the groundwater depending on their location. Players have three reusable drilling apparatuses which they use to drill for water samples. Players then must wait three minutes for the sample to return, meaning that students must develop sampling strategies in order to cover a large portion of the map. Because the GPS data is only accurate within 10 meters, there is some built-in error to the collected readings as well. Environmental Detectives contains a multimedia database of resources which students use to learn more about TCE, where TCE is found on campus, the health risks associated with exposure to TCE, and so on. Students access these resources by obtaining interviews from virtual university faculty and staff spread across campus. Because and there is not enough time to interview everyone or to drill more than a handful of wells, students must choose between collecting interviews, gathering background information, and drilling wells, adjusting and reprioritizing goals as new information becomes available. Using case study techniques, this paper describes findings gathered across five implementations of environmental detectives. Four of the cases were in university classes, the fifth was with secondary students. Researchers followed student groups, video taped actions, took field notes on emerging themes, and interviewed students in semi-structured interviews. Further explanation of the methodology and more detailed findings are provided in the full paper. Three major themes to emerge were navigation and space in augmented realities, the role of the physical environment in problem solving, and problem-solving patterns across gender lines. Students learned to navigate the augmented reality space with ease and enjoyed meeting virtual characters. Students in the college cases quickly took ownership over their problems, developing strong connections between the problem, physical space, and the simulated data provided on their PDAs. Students perceived the problem as authentic and legitimate and marshaled resources to develop solutions. In cases where students were less familiar with the environment, students had less emotional investment in the problem and found the problem to be more contrived. Students had different ways of solving problems. Some students, particularly men responded by drilling samples at the exclusion of interviewing experts or accessing library resources. Other students, particularly women, interviewed experts and accessed library resources before sampling. When groups had students embodying both these points of view, productive argumentation occurred; students debated the worth of ideas and developed plans encompassing both views. The paper concludes by exploring implications for the design of learning environments and next-generation augmented reality games.