This article identifies and discusses factors that influence the sample size necessary for longitudinal evaluations. These factors include the nature of the evaluation questions, nature of available comparison groups, expected consistency of the treatment in different sites, expected effect size, expected attrition rate, significance level for statistical tests, and the desired power of the statistical tests. Evidencefromfield-based evaluations is used in discussions of the relationships between these factors and sample size and the implications of those relationships for the evaluation planner.