One of the key components of any school choice system is the information given to parents as the basis for choice. We develop and implement a framework for determining the optimal performance metrics to help parents choose a school. This approach combines the three major critiques of the usefulness of performance tables into a natural, implementable metric. The best content for school performance tables is the statistic that best answers the question: “In which feasible choice school will a particular child achieve the highest exam score?” We implement this approach for 500,000 students in England for a range of performance measures. Using performance tables is strongly better than choosing at random: a child who attends the highest ex ante performing school within their choice set will ex post do better than the average outcome in their choice set twice as often as they will do worse.