It is a widely accepted idea that medications taken to relieve acute headache pain can paradoxically worsen headache if used too often. This type of secondary headache is referred to as medication overuse headache (MOH); previously used terms include rebound headache and drug-induced headache. In the absence of consensus about the duration of use, amount, and type of medication needed to cause MOH, the default position is conservative. A common recommendation is to limit treatment to no more than 10 or 15 days per month (depending on medication type) to prevent headache frequency progression. Medication withdrawal is often recommended as a first step in treatment of patients with very frequent headaches. Existing evidence, however, does not provide a strong basis for such causal claims about the relationship between medication use and frequent headache. Observational studies linking treatment patterns with headache frequency are by their nature confounded by indication. Medication withdrawal studies have mostly been uncontrolled and often have high dropout rates. Evaluation of this evidence suggests that only a minority of patients required to limit the use of symptomatic medication may benefit from treatment limitation. Similarly, only a minority of patients deemed to be overusing medications may benefit from withdrawal. These findings raise serious questions about the value of withholding or withdrawing symptom-relieving medications from people with frequent headaches solely to prevent or treat MOH. The benefits of doing so are smaller, and the harms larger, than currently recognized. The concept of MOH should be viewed with more skepticism. Until the evidence is better, we should avoid dogmatism about the use of symptomatic medication. Frequent use of symptom-relieving headache medications should be viewed more neutrally, as an indicator of poorly controlled headaches, and not invariably a cause.