Although most irritable bowel syndrome (IBS) patients are managed in primary care centers, most trials are performed in the hospital setting. A successful recruitment strategy is important for trial completeness and its external validity. To this end, it was important to assess the attrition rates of patients identified at primary care centers and hospitals in this phase II trial conducted in the outpatient clinics of 16 hospitals. IBS patients were identified through review of the centers' files (prescreening). After a 2-week single-blind placebo screening phase, the patients were randomized to receive an investigational drug or a matched placebo for 24 weeks (dose-ranging study). Thereafter, the patients were invited to participate in a 24-week, double-blind extension study. The attrition rates among patients identified at hospitals (group A) and at primary care centers (group B) were compared in each study phase and during the 50-week period by bivariate and multivariate regression analyses. Group A and B patients were identified in 13 hospitals and 51 primary care centers, respectively. Of 1,001 prescreened patients, 302 started the screening phase with attrition rates of 35% (of 132 patients) and 25% (of 170) for groups A and B, respectively (p = 0.054). The attrition rate during the double-blind phase was 14%. Of the 184 patients who completed the dose-ranging study, 39 (group A: 32%; group B: 14%; p = 0.005) did not wish to participate in the extension study. The attrition rate in the extension study was 15%. The overall attrition rates during the 50-week period were 67% and 53% (p = 0.016) for groups A and B, respectively. The multivariate regression analysis showed that the screening phases (p = 0.000) and unwillingness to participate in the extension study (p = 0.007) had the highest impact on attrition rates. We conclude that referring patients identified in primary care centers to hospitals seems appropriate to ensure potentially eligible patients for IBS trials. Patients identified in primary care centers are more likely than those identified in hospitals to participate in a 24-week extension study, which may be due to their positive feelings about being treated in a hospital rather than being referred back to their original primary care center. This strategy may be considered in future trials since, with a reasonably low attrition rate, it would enhance the external validity of the results obtained.