Clinical trials of therapies for acute migraine attacks have evolved over the years from open-label, small observational studies to highly structured randomised, controlled trials. The International Headache Society Committee on Clinical Trials in Migraine developed a tool to guide in designing scientifically sound trials. The proof of effect is best achieved in a clinical trial with:clearly defined objectives;a well-characterised study population, identified using well-validated diagnostic tools;proper randomisation and blinding;inclusion of a placebo arm, with proper balancing of patients receiving placebo and those receiving active drug;adequate study power; andappropriate statistical methods.Both parallel and crossover studies may be suitable in clinical trials of antimigraine agents, although the latter are a better choice in patient preference and bioequivalence studies.Although various efficacy measures are used to assess treatment effect, the 2-hour pain free rate (total resolution of pain within 2 hours after an initial moderate to severe headache) is preferred because it is clinically relevant and is relatively ‘placebo-insensitive’. Various migraine surveys have indicated that a rapid onset of therapeutic effect is a highly desirable attribute of an antimigraine drug. Therefore, accurate measurements of treatment effect before 2 hours are becoming increasingly emphasised.Consistency of effect across multiple attacks adds to the understanding of the therapeutic efficacy of a test drug. Finally, preference and satisfaction studies allow us to assess patients' global impression of a particular treatment, weighing the positive effects on pain and associated symptoms of migraine against potential adverse effects.