Article abstract Migraine is a heterogeneous condition that causes symptoms that vary both among individuals and within individuals from attack to attack. We examined and reviewed several important lessons on the diagnosis of migraine learned from the distribution of headache types and patterns of treatment response in the Spectrum Study, including recruitment and diagnostic issues. The accuracy of an initial diagnosis, assigned by a clinician in the context of a clinical trial, was compared with the results of a final diagnosis, assigned by a neurologist, reviewing the initial evaluation as well as headache diaries for up to 10 attacks. Several lessons can be learned from the Spectrum Study. Recruitment difficulties teach us that disabling tension-type headache is difficult to find, suggesting that it is rare. Examination of the final diagnosis given after diary evaluations suggests that a diagnosis of migraine can usually be confirmed for patients with disabling headache. After reclassification of the final sample of 432 subjects, 24/75 (32%) patients initially clinically classified as having disabling episodic tension-type headache proved to have migraine or migrainous headache after a diary review. Among study participants, 90% of subjects with disabling headache (HIMQ score >250) had a migraine-related disorder. Treatment response suggests that, in migraineurs, tension-type headaches may have a pathophysiology similar to that of migraine. The diary data show that mild headaches in patients with disabling migraine often evolve into full-blown migraine. The Spectrum Study supports the view that, for patients with disabling episodic headache, migraine is often the correct diagnosis. In clinical practice, the suspicion of migraine should be high for patients experiencing episodic disabling headache. Assessment of headache-related disability may assist practitioners in making a diagnosis of migraine.