This article describes the development and validation of two forms of the Geometry Assessments for Secondary Teachers (GAST), which were designed to assess teachers' knowledge for teaching geometry. Both forms were developed by teams of mathematicians, mathematics educators, psychometricians, and secondary classroom geometry teachers. Predictive validity for the GAST assessment was explored by observing and testing 157 teachers as well as administering pre- and post-tests to 3,698 students. The reliability coefficient for both GAST assessment forms was acceptable (r = 0.79). GAST assessment scores explained a statistically significant but small amount of the variance of student scores, demonstrating an effect that was greater than the number of years of teaching experience but smaller than the effect of having an advanced degree.