The extent to which the established variability in colonic motility recordings is due to differences in recording techniques is not known. The aim of this study was to compare the ability of two intraluminal recording devices (perfused tube and tube mounted strain gauge) to record colonic motor activity against a reference device (serosal strain gauge). In six anesthetized dogs an intracolonic probe was positioned such that the component perfused tubes and tube mounted strain gauges were approximated to identical strain gauges mounted on the serosa. Contractions were induced by field stimulation and intraarterial injections of acetylcholine and carbachol. While both intraluminal devices demonstrated limitations in the detection of phasic and tonic motor events, perfused tubes detected a significantly greater proportion of tonic and phasic contractions than did strain gauges (P less than 0.001). Intraluminal strain gauges misrepresented 50% of tonic contractions (confirmed visually and by serosal strain gauges) as waveforms with negative polarity. This was not seen in recordings from perfused tubes. Perfused tubes represented tonic contractions as biphasic or bifid waveforms significantly less frequently than strain gauges (P less than 0.05). Radial asymmetry of colonic contractions is likely to account for these observed discrepancies. Recorded motility patterns are influenced by different recording techniques, and these differences are a source of variability in recorded patterns of colonic motor activity.