Objective: To compare the effectiveness of manipulative therapy, physiotherapy, treatment by the general practitioner, and placebo therapy in patients with persistent non-specific back and neck complaints.

Design: Randomised clinical trial.

Setting: Primary health care in the Netherlands.

Patients: 256 patients with non-specific back and neck complaints of at least six weeks' duration who had not received physiotherapy or manipulative therapy in the past two years.

Interventions: At the discretion of the manipulative therapists, physiotherapists, and general practitioners. Physiotherapy consisted of exercises, massage, and physical therapy (heat, electrotherapy, ultrasound, shortwave diathermy). Manipulative therapy consisted of manipulation and mobilisation of the spine. Treatment by general practitioners consisted of drugs (for example, analgesics), advice about posture, home exercises, and (bed)rest. Placebo treatment consisted of detuned shortwave diathermy (10 minutes) and detuned ultrasound (10 minutes).

Main outcome measures: Changes in severity of the main complaint and limitation of physical functioning measured on 10 point scales by a blinded research assistant and global perceived effect measured on a 6 point scale by the patients.

Results: Many patients in the general practitioner and placebo groups received other treatment during follow up. Improvement in the main complaint was larger with manipulative therapy (4.5) than with physiotherapy (3.8) after 12 months' follow up (difference 0.9; 95% confidence interval 0.1 to 1.7). Manipulative therapy also gave larger improvements in physical functioning (difference 0.6; -0.1 to 1.3). The global perceived effect after six and 12 months' follow up was similar for both treatments.

Conclusions: Manipulative therapy and physiotherapy are better than general practitioner and placebo treatment. Furthermore, manipulative therapy is slightly better than physiotherapy after 12 months.