This empirical study was conducted to explore both the influence of assessment on teachers' classroom practice during an assessment driven educational reform, and the principle underlying assessment driven instruction, i.e., "what you test is what you get." Factors targeted at the level of assessment were performance-based or objective paper-and-pencil methods. Factors targeted at the level of the teacher were demographic variables, teachers' opinions of assessment, the assessment method, teachers knowledge of assessment, and teacher efficacy. In the Fall of 1992, self-report data were collected from 117 regular classroom teachers. According to the data analysis, the research did not support the "what you test is what you get" assumption. Rather, it was determined that teachers' classroom practices were influenced by methods of assessment, teachers generally taught toward the method of state-wide assessment most congruent with their existing classroom practices, and the feasibility of assessment driving teachers' classroom practices varies as a function of several factors endogenous to the teacher as well as to the assessment. Implications for educational policy-makers who plan assessment-driven reforms and statistical tables complete the document. Contains 47 references. (LL)