This article summarizes a 2000 study of school garden programs and their variation and the impact of such variation on 427 third-grade students' sense of responsibility and attitudes toward science and the environment. A teacher questionnaire was developed to gain insight into how teachers use school gardens with their students and in their curriculum. The information gathered from 28 third-grade teachers was used to develop a classification framework or typology of garden types that served as the independent variable of analysis. Data on school garden program variation was simplified into a typology based on intensity, measured by the number of garden-related activities students participated in prior to and while in the garden (high, medium, and low), and the form of school gardens (flower, vegetable, or combination flower/vegetable), resulting in nine garden types. Analysis of covariance tests were used to determine if there were significant differences in the nine types of school gardens. Significant differences were found in the school garden types and students' attitudes toward science and attitudes toward the usefulness of science study. Although there were no significant differences in school garden types and students' responsibility scores and environmental attitudes, scores for each of these elements were very high (indicating a sense of responsibility and a positive environmental attitude) with little variation. (Contains 6 tables.)