We studied the process and outcomes of physical therapist management of 107 patients with low back pain at a walk-in clinic. Sixty-seven patients with low back pain concurrently were assigned randomly to internists at the clinic. Physicians and physical therapists recorded baseline clinical data and management plans on standard check lists. Physical therapists used a validated algorithm that directed diagnostic evaluations and physician consultations, but all other physical therapist treatment decisions were unconstrained. Patients in the physical therapist and physician groups did not differ significantly. Physical therapists referred more patients to the physical therapy department than did physicians but recommended muscle relaxants, prescription analgesics, and bed rest less frequently. The occurrence of new symptoms, duration of symptoms, and duration of activity limitations were similar between the physician and physical therapy patient groups at a one-month follow-up examination. Physical therapist-managed patients expressed greater satisfaction than physician-managed patients with several aspects of their care. The percentage of functional improvement for highly dysfunctional patients was significantly greater for the physical therapist-managed patients than for the physician-managed patients. The implications of a physical therapist first-contact care program for health service organizations, health care policy, physical therapist training, and credentialing are discussed.