The investigators will perform a cluster randomised controlled study of Brief intervention (BI) for medication-overuse headache (MOH) versus business as usual. GPs will be trained to perform a structured brief intervention after identifying patients with probable MOH using the severity of dependence scale. The control arm will include patients of GPs who have not been trained in BI. Patients will be recruited by prior short postal screening of patients listed on the GPs patient lists.

The hypothesis is that BI will lead to improvement of medication-overuse and chronic headache as compared to no BI.

Main outcomes are:

• number of medication days per month
• number of headache days per month
• headache index

123 cases in all included (75 in intervention/placebo control arms and 48 in non-intervention Control arms) 19 did not meet predefined inclusion criteria for the different arms (5 and 8 in intervention arm and placebo arm respectively, 4 and 2 in chronic headache Control and population Control respectively)

1 case was lost to follow up in each of the 4 Groups prior to first assessment. Thus 60 cases remained in intervention/placebo Control arms (24 in intervention and 36 in placebo Control) and 40 in non intervention Control arms (15 chronic headache Controls, 25 population Controls).

Inclusion Criteria:

• 18-50 years of age
• Screening positive for possible chronic headache (> 10 headache days per month) (not for population control arm)
• Screening positive for possible medication-overuse (> 15 medication days per month) (not for chronic headache control arm)

Exclusion Criteria:

• Other complicating pain with medication treatment