This essay argues that the secondary analysis of existing data sets provides an important opportunity for researchers concerned with science education. The traditional arguments for the secondary analysis of large data sets-quality, costs, and the “science” of multiple analyses of the same data-are reviewed. The increasing number and quality of databases available, the introduction of new statistical technologies, and the severe reduction of federal support for science education are three new forces fostering secondary analysis in science education. Some cautions for secondary analysts are shared and three model secondary analyses are described. The essay concludes that the forces favoring secondary analysis in science education are increasing and that the completion of some secondary analysis prior to new data collection may become the normative expectation in the science education community.