The purpose of this work was to study the impact of a patient-self-completed history questionnaire upon the quality of the information in the medical record, resource utilization, patient satisfaction, and physician satisfaction. A controlled trial was performed in a primary care clinic of a public supported, urban, university hospital. The patients were mainly poor, minority, urban individuals visiting the clinic for their first primary care visit. The providers were attending physicians, nurse practitioners, and medical residents associated with the Department of Medicine teaching program. A self-administered history questionnaire was given to approximately half the new patients in a randomized, prospective manner. The physicians completed an assessment form on all new patients estimating the time spent, complexity of the case, and listing studies requested. If a history questionnaire was completed by the patient, the physician completed an assessment irrespective of whether it helped or hindered the evaluation of the patient. After the visit, nurses called the patient by telephone and asked questions concerning the patient's satisfaction. The patients' records were reviewed for the initial history papers. There were 127 patients receiving the history questionnaire and 114 controls. The presence of the history questionnaire in the medical record added to both the quantity and the quality of the information present. Of the provider responses, 87% (95% confidence interval: 78-93%) wanted all new patients to complete the questionnaire. Of the patients who received the questionnaire, all who were asked felt that all their issues were addressed, compared with only 83% in the control group (P = .015). No increases in physician time or in utilization of medical resources were found. The implementation of a self-administered history questionnaire in an urban primary care clinic resulted in improved chart quality and improved satisfaction of physicians and patients.