Evaluation of complementary feeding programs is needed to enhance knowledge on what works, to document responsible use of resources, and for advocacy. Evaluation is done during program conceptualization and design, implementation, and determination of effectiveness. This paper explains the role of evaluation in the advancement of complementary feeding programs, presenting concepts and methods and illustrating them through examples. Planning and investments for evaluations should occur from the beginning of the project life cycle. Essential to evaluation is articulation of a program theory on how change would occur and what program actions are required for change. Analysis of program impact pathways makes explicit the dynamic connections in the program theory and accounts for contextual factors that could influence program effectiveness. Evaluating implementation functioning is done through addressing questions about needs, coverage, provision, and utilization using information obtained from process evaluation, operations research, and monitoring. Evaluating effectiveness is done through assessing impact, efficiency, coverage, process, and causality. Plausibility designs ask whether the program seemed to have an effect above and beyond external influences, often using a nonrandomized control group and baseline and end line measures. Probability designs ask whether there was an effect using a randomized control group. Evaluations may not be able to use randomization, particularly for programs implemented at a large scale. Plausibility designs, innovative designs, or innovative combinations of designs sometimes are best able to provide useful information. Further work is needed to develop practical designs for evaluation of large-scale country programs on complementary feeding.