IntroductionPosterolateral spinal fusion has become an evidence-based treatment of chronic low back pain caused by spondylolisthesis or spinal instability [12,26]. Although widely used, the beneficial effect on the clinical outcome of adding pedicle-screw fixation to the posterolateral spinal fusion has not yet been demonstrated [11,13,25,38]. All studies have shown reduction in pain and disability, but few have investigated the severity and localisation of residual pain several years after the procedure.Abstract Pain drawings have been used in spine surgery for diagnostic use and psychological evaluation of fusion candidates; they have rarely been used to evaluate pain status after spinal fusion. This study is a 5-year follow-up on a randomised clinical trial assigning patients to posterolateral spinal fusion with or without pedicle screw instrumentation. Patients were mailed a pain drawing and questionnaires including questions regarding work, social status, smoking status, the Dallas Pain Questionnaire (DPQ), and the Low Back Pain Rating Scale (LBPRS). Pain drawings were scored using a visual inspection method and a surfacebased point scoring and evaluated for the presence of donor site pain. Pain drawings from 109 patients (87% of the initially included patients), 56 men and 53 women, mean age at followup 51 years, were analysed. Fiftythree patients had undergone an instrumented fusion and 56 a non-instrumented fusion. Some presence of low back pain was marked by 79% and leg pain by 69%. Sixty-two percent of the pain drawings were classified as "organic" and 38% as "nonorganic". There was no difference between the instrumented and the uninstrumented group. DPQ and LBPRS scores were higher in the non-organic group (P=0.007). Using the point scoring, no difference between the instrumented and the uninstrumented group was seen. The results of the point scoring were found to correlate with the DPQ and LBPRS scores (P=0.001). Working patients (39%) had significantly better scores than the rest. Ten percent of the patients had donor site pain. Twenty percent of spinal fusion patients are totally pain free at 5-year follow-up. Ten percent still experience donor site pain. In general, instrumentation does not affect the amount and localisation of pain 5 years after lumbar spinal fusion surgery. The pain drawing seems to be a valuable tool when following spinal fusion patients, but its use as prognostic marker in connection with fusion surgery needs further investigation.