There is no comprehensive database of U.S. district-level test scores that is comparable across states. We describe and evaluate a method for constructing such a database. First, we estimate linear, reliability-adjusted linking transformations from state test score scales to the scale of the National Assessment of Educational Progress (NAEP). We then develop and implement direct and indirect validation checks for linking assumptions. We conclude that the linking method is accurate enough to be used in analyses of national variation in district achievement, but that the small amount of linking error in the methods renders fine-grained distinctions among districts in different states invalid. Finally, we describe several different methods of scaling and pooling the linked scores to support a range of secondary analyses and interpretations.