This study investigated the effects of using Geographic Information Systems (GIS) to improve middle school students? and their teachers' understanding of environmental content and GIS in a constructivist classroom. Constructivism provided the theoretical framework with Bonnstetter's inquiry evolution and Swartz's skillful problem solving as the conceptual framework for designing these GIS units and interpreting the educational results. Teachers from nine schools in five counties in North Carolina attended a one-week workshop and fall follow-up session, where they learned how to teach the online Mapping Our School Site (www.ncsu.edu/scilink/studysite) and CITYgreen GIS inquiry-based problem-solving units. Two years after the workshop, one science instructor and one media specialist from the workshop taught the six week Mapping Our School Site (MOSS) unit in the fall and one biotechnology elective teacher taught the MOSS unit in the fall and the CITYgreen GIS unit in the spring. The students in the MOSS experimental group (n=131) and the CITYgreen GIS comparison group (n=33) were compared for differences in understanding of environmental content. Following the teaching of environmental science with GIS, other factors were investigated such as students' spatial abilities, experiences, and learning preferences. Teachers and students completed the online Learning Styles Inventory (LSI), Spatial Experience Survey (SES), and the Purdue Spatial Visualization Test: Rotations (PSVT:R). Using ANOVAs, a quiz, unit test, analyses, written conclusions and lab reports indicated that the CITYgreen GIS group learned the environmental content better than the MOSS group. The MOSS group understood how to design experiments and to use GIS better to analyze problem questions. Both groups improved in problem identification and problem solving, data accuracy, and hypothesis testing. An ANOVA was used to compare the spatial reasoning score to learning style as reported on the LSI, and factors on the SES such as handedness, academic courses (geometry, shop class, 3-D courses), non-academic activities (toys, computer games), and sports activities. Males scored higher than females on the spatial reasoning test, the more computer games played the higher the score, and the fewer shop classes taken the higher the score. Earlier research asserts that qualified teachers who are confident in teaching with GIS are key to the success of the curricula. The effectiveness of the MOSS/CITYgreen GIS teacher workshop was also investigated. Results indicated that 75% of the teachers' integrated GIS into classroom instruction two years after the GIS workshop. Even though teaching experience was negatively related to spatial reasoning test scores, implementation of GIS by teachers in the workshop was not influenced by years of teaching experience. The results indicate that GIS can be universally used for classroom instruction. Using qualitative analyses, the results of the quantitative analysis were further evaluated. This approach indicated that the students learned more than was reported by the unit test. Even though the CITYgreen GIS students learned more environmental content than the MOSS students, the MOSS students were better able to analyze problem questions using GIS. The findings of this study indicate that using GIS may aid students in constructing concepts and promoting understanding of environmental content, problem solving, experimental design and data analysis, and communicating findings to others. Using GIS in classroom instruction may be a way of incorporating spatial learning in schools.