Hypothetical comparative effectiveness studies of two migraine drugs offer an opportunity to analyze the legal issues that manufacturers face in conveying to practicing physicians information from such research about prescription drugs. This information, derived from observational studies, supposedly cannot be communicated to clinicians because it does not constitute “substantial evidence” as required by law. In fact, however, the emerging effectiveness data can be included in manufacturers’ communications to physicians about drugs’ clinical utility through various existing channels defined by other law. This article argues that Congress should further clarify the circumstances under which such communications can occur to comply with First Amendment requirements that speech constraints be narrowly drawn, known in advance, and precise. Otherwise, the Food and Drug Administration risks losing First Amendment arguments in cases involving the agency’s regulation of the communications—or “speech”—of drug manufacturers.