Little attention has been paid to methodological aspects in the recording of gastrointestinal symptoms. We compared a new fixed-point scale for the self-recording of pain intensity with steps operationally linked to behavioral events and with additional monitoring of pain duration--termed the "duration-intensity-behavior scale" (DIBS)--with the visual analogue scale (VAS) in 32 patients with nonulcer dyspepsia. After randomization, the patients either recorded pain intensity (VAS) or pain intensity and duration (DIBS) four times daily during a 4-week period for 1 preliminary week without medication, followed by 3 weeks of antacid treatment. For both scales there was a high degree of compliance, and they seemed equally reliable and sensitive to changes in pain experience. Since DIBS yields more clinically useful information, this scale appears to be preferable for the monitoring of gastrointestinal pain.