Abstract Although there has been a ‘dramatic drift’ towards experimentation in political science, the methodology remains scarce in public management research. This article considers the potential for the experimental method for public management. It discusses the benefits and costs of an experimental design. It identifies three barriers to the use of experiments distinctive to public management; a stress on realism and practical solutions; a focus on organizations rather than individuals; and ethical and logistical challenges. It re-evaluates these barriers in the light of recent experiments, arguing that experimental approaches should now be added to the toolkit of public management research.