The prevalence of irritable bowel syndrome (IBS) in psychiatric practice was studied in 41 consecutive psychiatric outpatients. Different criteria for IBS were applied to the data set to determine the effects on the rates of IBS obtained. Depending on an option in the Rome criteria, IBS rates varied from 13% using the 'and' requirement for combining abdominal pain and altered bowel function symptoms, to 41% using the 'and/or' option described in the formal definition statement in 1990. The resultant prevalence rates of IBS varied greatly according to which published criteria were applied, with a maximum of 71%. This wide variation in rates depending on the criteria underscores the critical importance of standardizing diagnostic research criteria for IBS. An exemplary model of empirically based validation has been developed for psychiatric disorder criteria which, like IBS, are symptom- based and lack physiological determinants. Validated diagnostic criteria for IBS await similar study. (C) 2000 by W.B. Saunders Company.