Chronic use of symptomatic headache medication is believed to be a risk factor for drug-induced chronic headache. This study reports the frequency of chronic use of symptomatic headache medications implicated in drug-induced chronic headache among primary-care headache (PCH) patients and identifies predictors of chronic/frequent use. The design uses a 2-year cohort study of a sample of PCH patients in Group Health Cooperative of Puget Sound, a large health maintenance organization in Seattle, WA. Among 779 PCH patients (aged 18-74 years) interviewed at baseline, 662 (85%) completed both 1- and 2-year follow-up interviews. This study estimates the percent of PCH patients reporting: (1) frequent use of over-the-counter, non-steroidal anti-inflammatory, opioid, sedative-hypnotic, and ergot medications (defined as 14 or more days of use for each class in the prior month), (2) chronic/frequent use of these medications, defined as surpassing criteria for frequent use on at least 2 of the 3 study interviews, and (3) chronic/frequent polypharmacy, defined as chronic/frequent use of 2 or more classes of headache medication. Twenty-one percent of PCH patients were chronic/frequent users of symptomatic headache medications, while 2.6% met study criteria for chronic/frequent polypharmacy. Chronic/frequent use of over-the-counter medications (15.9%) was twice as common as chronic/frequent use of prescription medications (7.7%). Headache days at baseline was the strongest predictor of chronic/frequent medication use. After controlling for baseline persistence and severity, older patients were more likely to be chronic/frequent users of headache medications. The age effect was explained by chronic/frequent use of over-the-counter medications.(ABSTRACT TRUNCATED AT 250 WORDS)