The majority of states include value-added models (VAMs) or some other measure of student academic growth as a component of their teacher evaluation systems. However, there is considerable disagreement among researchers about whether states and school districts should use student growth measures to make high-stakes personnel decisions. Many researchers have concluded that VAMs and other student growth models are not appropriate as the primary measure for evaluating individual teachers. Others maintain that VAMs provide important and useful information on the effects that teachers have on their students' achievement. Studies on the consequences of using VAM scores in teacher evaluation systems are still accumulating. However, evidence has begun to emerge that teachers' VAM scores may depend more on the students they teach and the schools where they work than on the effectiveness of their teaching. This Information Capsule reviews the problems researchers have documented with using VAMs as measures of teacher effectiveness. For example, teachers' VAM scores vary depending on the students in their class, the school where they work, the specific achievement test used, and the statistical model used in the calculations; a teacher's VAM score can vary substantially from year to year; VAM scores are not highly correlated with other measures of teacher effectiveness; and it is difficult to isolate the impact of a single teacher on students' academic growth. Researchers have urged caution when basing teacher evaluations on VAM scores. Their recommendations for calculating and interpreting VAM scores, summarized in this Information Capsule, include using VAM scores as only one component of a comprehensive teacher evaluation system, using multiple years of data when calculating VAM scores, and calculating VAM scores only in grades and subjects where there are highly reliable and valid assessments that are comparable over time.