When the task of evaluating student learning is carefully considered, two major problems emerge. One is the gathering of the most appropriate and precise evidence possible about the learning. The other is the setting of performance standards against which this evidence may be weighed and the adequacy of each student's learning judged. This paper has focused on the problem of setting performance standards for use in strategies for mastery learning. The paper began with the argument that a key variable in the design of these strategies are the mastery performance standards which students are helped to attain throughout their instruction. It was pointed out that presently there are no procedures for setting such standards. Next, an attempt was made to formulate one such procedure. The approach developed utilizes students' future learning, i.e., their scores on a set of desired, end-of-instruction learning outcomes, as a criterion for determining the mastery performance level which students must attain at any stage in their instruction. Finally, the paper reported an experiment designed to explore the feasibility of the approach proposed. The experiment was designed to test the assumption that the performance standard which a student attains over each segment of his instruction has important implications for his realization of the desired, end-of-instruction learning outcomes. In general, the experiment's results confirmed the assumption tested. (Author/CK)