OBJECTIVES: The definition of chronic migraine (CM), as opposed to episodic migraine, has evolved over 20 years of clinical study. This systematic review summarized population-based studies reporting CM prevalence and examined how variations in definitions affected estimates. METHODS: We conducted a systematic search of PubMed, Cochrane, and EMBASE. Search terms included chronic migraine, transformed migraine, chronic headache, and prevalence. We included population-based studies in adults and mixed adults/adolescents that estimated CM prevalence (or provided information to calculate estimates). We defined chronicity as a minimum headache frequency of 10 per month or reported "daily" headache. Where equivalent, we combined diagnostic criteria and definitions. Findings were qualitatively summarized. RESULTS: Sixteen publications representing 12 unique studies were identified. Diagnostic criteria were heterogeneous and included definitions based on International Classification for Headache Disorders (ICHD) guidelines, Silberstein-Lipton criteria, and various investigator-defined frequency-based classifications. Thus, definitions varied from relatively strict criteria (≥15 days/month of migraine) to more liberal criteria (history of migraine and ≥15 days/month of headache). Prevalence of CM was 0%-5.1%, with estimates typically in the range of 1.4%-2.2%. Seven studies used criteria of history of migraine and ≥15 days/month of headache (or equivalent), with prevalence of 0.9%-5.1%. Three studies used criteria of ≥15 days/month of migraine (or equivalent), with prevalence of 0%-0.7%. Two studies stated that the condition was transformed migraine or chronic migraine (without specific criteria), with prevalence of 1.6% and 4.1%. Prevalence varied by WHO region and gender. CONCLUSIONS: CM prevalence estimates are influenced by specific definitions employed. Using the strictest criteria, prevalence was well under 1%; with a less restrictive definition, prevalence was higher, 1%-5%. With these variations, it is difficult to compare results across regions and explore temporal trends. Future studies on CM would benefit from an ICHD consensus diagnosis that is clinically appropriate and operational in large-scale epidemiological studies.