Abstract Background Treatment of headache disorders is not always optimal. Patients are treated in multiple ways, and the lack of scientific arguments for referral and the insufficient implementation of guidelines result in unclear treatment strategies. The coexistence of headache and neck pain can lead to the referral to a musculoskeletal physiotherapist. This treatment can only be successful if an underlying cervical segmental dysfunction is present. In such cases a physical treatment can be a valuable option that should be considered. The aim of this study is to identify prognostic therapeutic patient characteristics and to increase the number of correct physiotherapy referrals. Methods/design This trial is designed to identify patient characteristics which can influence the prognosis of the patient. Patients with recurrent headache and co-existent neck pain are recruited via a multicenter setup. After screening for eligibility, subjects are tested at baseline and randomly allocated to one of two treatment groups. Testing includes the administering of questionnaires (a Headache Diagnosis Questionnaire, Headache Inventory List and the Headache Impact Test (HIT-6)) and physical tests (Thermal Stimuli, Manual Cervical Spine Examination and Pressure Algometry). Treatment groups are a usual care group (UC) administered by the General Practitioner (GP) and a usual care plus musculoskeletal physiotherapy treatment group (UCMT). UC is based on the Dutch GP Guideline for Headache. UCMT consists of the UC plus a combination of exercises and spinal cervical mobilisations. Follow-up measurements consist of the completion of the Headache Inventory List, the HIT-6 and scoring of the global perceived effect (GPE). The latter allowing the distinction between responders (positive effect) and non-responders (no effect or worse). Logistic regression analysis will be used to identify the specific patient characteristics of the responders and the non-responders. The additional value of the musculoskeletal physiotherapy will be examined. Follow-up measurements up to 52 weeks are scheduled. Discussion This trial aims to identify prognostic patient characteristics, in order to supply a useful diagnostic tool for all health care workers, dealing with headache sufferers.