Questionnaires are widely used instruments to monitor gastrointestinal (GI) symptoms. However, few of these questionnaires have been formally evaluated. We sought to evaluate our GI symptoms questionnaire in terms of clarity and reproducibility. Primary care patients referred for open access Helicobacter pylori urea breath testing reported GI symptoms (type+severity) and demographic information by written questionnaire. In an interview, patients gave a personal description of the meaning of the GI symptoms on the questionnaire. Patients' descriptions of GI symptoms were compared with current definitions. Symptom severity scores were compared before and after, interview versus questionnaire. Of the 45 patients included, 19 (42%) described all symptoms correctly, whereas 17 (38%) described one symptom incorrectly. None of the patients made more than three mistakes. Regurgitation was the most common incorrectly described symptom (16 patients [36%]), whereas the other individual symptoms were well explained. Symptom severities before the interview, after the interview and reported by questionnaire (mean value+/-SEM) were 2.1 +/- 0.2, 2.1 +/- 0.2, and 1.5 +/- 0.2 points on a 7-point Likert scale (0-6), respectively. Mean severity reported by interview (95% CI) was 1.4 (1.3-1.5) times higher than reported by questionnaire (P < .05). In conclusion, the GI symptom questionnaire is understandable and has good reproducibility for measuring the presence of GI symptoms, although symptom severity is consistently rated higher when reported by interview.