In general, the interaction between statistical and policy analysis has the potential results of providing the legal policymaker with premises which may be used for deductive prediction or policy optimization; and by being sensitive to policy analysis, the statistician can determine what statistical procedures are most useful to policymakers and adjust inference, prediction, and causation models accordingly. This article examines five areas in which statisticians and legal policymakers can achieve this synergistic relationship: (1) predicting the effects of smaller juries, (2) determining an optimal percentage of defendants to hold prior to trial, (3) identifying racial differences in sentencing, (4) predicting personal injury case outcomes, and (5) allocating anticrime dollars across places and activities. Some conclusions are that the statistician must recognize, first, that causal analysis may mean one thing when used to calculate the theoretical relationships between a variety of variables and something different when designed to serve as part of an optimum allocation policy; and second, prediction analysis may require different methods when policy-relevant outcomes are involved, since such outcomes facilitate assessing the costs of mispredictions. Especially important for policy analysts are statistical inputs such as surveys of values to be maximized. Policy analysts might also find predictive or causal regression analysis useful in predicting relations between alternative policies and goals. National meetings of political-legal and statistical organizations and interdisciplinary policy studies programs are important steps in increasing the fruits of statistical analysis for legal policymaking. Illustrative tabular and graphic data and 10 footnotes are provided.