Abstract
At 45 general-practice surgery sessions 200 patients in whom no definite diagnosis could be made were randomly selected for one of two procedures. Either they were given a symptomatic diagnosis and medications, or they were told that they had no evidence of disease and therefore they required no treatment. No difference in outcome was found between these two methods as judged by the return or not of the patient within one month and his statement that he did or did not get better.