What evidence should guidelines take note of?

The Guidelines Challenge Conference on which this special issue builds asked as the first of its “further relevant questions”: “How do we incorporate more types of causally relevant information in guidelines?” This paper first supports the presupposition of this question—that we need further kinds of evidence—by pointing out that the randomized controlled trial, touted as the best source of evidence on effectiveness, can do so little for us. Second, it outlines a number of other good ways to learn what will work that the medical community, and much of the public health community, is not making much use of.