The article begins with a presentation of reasons for the minimal influence of homework research, and educational research in general, on policy and practice. We propose that the practical use of research is hampered by the complexity of real-world settings, by flawed research designs, and by the probabilistic nature of research outcomes. Systematic research syntheses are offered as providing one way to overcome these difficulties. This special issue is a case example. To set the stage for the articles and discussion that follow, a definition for homework is offered along with a brief history of public attitudes toward homework and a listing of positive and negative effects attributed to homework. Then, the results of a research synthesis and a survey study are described that provide evidence for answering a complex and controversial question: How much time should students spend on homework each night? Little association is found between the amount of homework young students complete and achievement. The association grows progressively stronger for older groups of students. Other research suggests that young children have limited ability to keep their attention focused and have not learned good study skills. Two examples are provided showing how the research results can be used to evaluate the appropriateness of recommendations for policy and practice.