Pain disorders that are primarily associated with psychological factors are of great clinical concern, but they are difficult to study because of the inability to make valid or reliable diagnoses by structured interview alone. The authors confront this difficulty by using an injured subject population that had extensive psychiatric and medical evaluations. Those who developed somatoform pain disorder (SPD) were compared with a control group who did not. The SPD group had distinctive associated factors: more sites of pain, spread of pain beyond area of original injury, and substantially more opiate and benzodiazepine use. Compensation/litigation influenced symptoms more in the SPD group. Psychotherapists often supported the patient's viewpoint that the pain was physical and to be endured.