The objective is to present the design of a randomized clinical trial (RCT) on the effectiveness and cost effectiveness of a behavioural graded activity programme compared with manual therapy in patients with sub-acute neck pain. Sub-acute is defined as pain existing for 4-12 weeks. The behavioural graded activity programme is a time-contingent increase in activities from baseline towards pre-determined goals. Manual therapy consists mainly of specific spinal mobilization techniques and exercises. The primary outcomes are global perceived effect and functional status. Secondary outcomes are kinesiophobia, distress, coping, depression and somatization. The intensity and persistence of the pain and its interference with activities are also assessed. Direct and indirect costs are measured by means of cost diaries. Measurements take place at baseline and 6 and 12 weeks after randomization. To assess the long-term effect, measurements will also take place after 6 and 12 months. Finally some challenges are discussed concerning the use of a behavioural graded activity programme, manual therapy and outcomes.