BACKGROUND: There is evidence for an association between Modic type 1 and pain in patients with low back pain (LBP), but little knowledge about its effect on clinical outcomes. PURPOSE: (1) To assess the prevalence of Modic changes, (2) to determine if Modic changes influence the clinical course of LBP, and (3) to identify prognostic factors for recovery. STUDY DESIGN/SETTING: Prospective clinical cohort study, with a 1-year follow-up. The treatment consisted of brief intervention and instruction in stretching. PATIENT'S SAMPLE: Two hundred and sixty-nine patients with chronic low back pain. OUTCOME MEASURES: Socio-demographic variables, education, profession, self-reported measures, degenerative changes on MRI. METHODS: Linear growth model and Cox regression analysis. RESULTS: Five percent had a normal MRI, 14 and 50% had Modic 1 and Modic 2 changes. Modic changes were not significant covariates for the clinical course of pain, function or fear avoidance beliefs. Education was a strong prognostic factor for recovery. CONCLUSIONS: Modic changes did not influence the clinical course of back pain and were not prognostic factors for recovery. Education was strongly associated with recovery.