The proposed science court will not make decisions nor recommendations concerning public actions on the basis of the facts it presents. Its fact-finding is intended to enable nonscientists to make informed decisions that incorporate scientific value systems. In the science court procedure, advocates for the scientific facts at issue will present the factual bases for their positions before scientifically trained judges who have no involvement in the field in question as well as in the presence of expert adversaries. An analog of cross-examination will be used to exhibit the weaknesses in claimed scientific factual statements. Any statement or conclusion by the science court will be valid only until new evidence on the issue arises. The outcome of a science court session will be the determination of what is known and unknown about a particular scientific issue. This article first examines arguments in favor of the creation of a science court and then compares the truth-seeking devices of the scientific method and the legal system to assess their merits in assisting public policymakers faced with issues that involve scientific matters. After a discussion of the various models proposed for a science court, the article reviews the model proposed by Arthur Kantrowitz and defends it against criticism while suggesting refinements. The potential dangers of the science court are identified. Overall, the article concludes that an experimental science court is justified if for no other reason than to assess how it performs in practice. 33 footnotes