Background: Chronic gastrointestinal (GI) symptoms are believed to be common in the general population, but Australian data are lacking. A valid instrument is required to assess GI symptoms adequately and determine their prevalence in the community.

Aims: To test the feasibility, reliability and concurrent validity of a self-report Bowel Symptom Questionnaire (BSQ) as a measure of GI symptoms, and obtain preliminary data on the prevalence of symptoms in an Australian population-based sample.

Methods: Outpatients (n = 63), volunteers (n = 163) and a random sample (n = 99) of the Penrith population, Sydney, completed the BSQ. Feasibility was evaluated in 264 subjects. Reliability was measured by a test-retest procedure (n = 43), while concurrent validity was documented by comparing self-report data with an independent interview (n = 20). The response rate in the population mail survey was 68%. Prevalence data on bowel symptoms in the community sample (n = 99) were age and gender standardised to the Australian population.

Results: The majority of subjects found the BSQ easy to complete (97%) and understand (97%); 90% completed the questionnaire in half an hour or less. Reliability (median kappa 0.70, interquartile range 0.20) and concurrent validity (median kappa 0.79, interquartile range 0.26) of GI symptoms were both very acceptable. The internal consistency of all GI symptom scales was good (Cronbach's Alpha range 0.51-0.74). The prevalence of the irritable bowel syndrome (defined as abdominal pain and disturbed defaecation based on two or more of the Manning criteria) was 17.2% (95% CI: 10-25%).

Conclusions: The BSQ was well accepted and easy to understand; it provided reliable and valid data on GI symptoms and should prove useful in large scale epidemiological studies in Australia.