The 'Rome' criteria for irritable bowel syndrome (IBS) have evolved over 15 years with four published versions. The impact of these changes on community prevalence rates is not known. Study aims were to estimate the prevalence of IBS using the four Rome criteria and agreement between Rome II and previous criteria. Questionnaires were mailed to a random sample of Olmsted County, Minnesota residents in 1992. Age- and gender-adjusted prevalence estimates were calculated for Rome II (1999), Rome I (1992), Rome (1990), and Rome (1989) criteria. Per cent agreement and kappa values were calculated to assess agreement. Of 892 eligible subjects, 643 (72%) responded. The age- and gender-adjusted prevalence of IBS was 5.1% [95% confidence interval (CI): 3.1, 7.0], 6.8% (95% CI: 4.7, 8.9), 5.1% (95% CI: 3.2, 7.1) and 27.6% (95% CI: 23.6, 31.5), respectively. In comparison with Rome II criteria, per cent agreement and kappa values were 97.2% and 0.78 (95% CI: 0.69, 0.88), 96.4% and 0.68 (95% CI: 0.55, 0.80), and 79.0% and 0.29 (95% CI: 0.19, 0.40), respectively. Thus, although differences were seen with the older criteria, compared with the Rome I criteria, good agreement was seen and community prevalence estimates were similar with the Rome II criteria.