The aim of this study was to evaluate methodologic aspects of colonoscopic laser Doppler flowmetry. A Periflux PF1d flowmeter, set to 4 kHz/0.2 sec, with an endoscopic probe (PF 109) was used. In 20 patients, with a median age of 70 years and without colonic disease, flux was recorded at 10, 40, 30, 20, and again at 10 cm from the anal verge. A median of three repeated recordings were made at each level, to calculate average flux and spatial variation. Median flux was 158 (150-167) perfusion units, and the coefficient of variation of repeated recordings 0.14 (0.12-0.17). There was no regional variation, and no increase in flux at 10 cm from the start until the end of the procedure. Pressure of the probe against the bowel wall and severe distention significantly reduced the flux. The interference of light from the endoscopic light source on the flux could not be predicted. It differed with different light sources, and also with the length of probe coming out of the colonoscope--that is, the distance from the light to the measurement point. To avoid the problem, the light source should be turned off while recording.