Back-pain patients with onset in the preceding 1-10 days and comparable on a back examination were randomly assigned to traditional management (A regimen) and behavioral treatment methods (B regimen). Patients were compared at 6 weeks and 9-12 months on a set of "Sick/Well" scores derived from patient reported vocational status (V), health-care utilization (HCU), claimed impairment (CI), and pain drawings (D) and on two measures of activity level. No differences were found at 6 weeks, but at 9-12 months, A-group S's were more "sick." No A/B differences were found on activity-level measures. Group A S's showed significant increases in claimed impairment from preonset to follow-up, whereas Group B S's had returned at follow-up to preonset levels.