Ninety-two chronic low back pain patients were randomly allocated to two groups to evaluate the effectiveness of a back school compared with an exercise-only regimen according to specified outcome variables. The data from 78 patients with 7 years mean duration of symptoms was analyzed. Three assessments were made: before treatment and 6 and 16 weeks after treatment. Changes in patients' levels of pain, functional disability, and other related variables were compared in the two groups. Almost all variables showed an improvement at 6 weeks. At 16 weeks, functional disability and pain levels showed a significant difference. Back school patients continued to make an improvement. This method of managing low back pain makes maximal use of limited resources and appears to be effective, especially in the longer term.