An international working team of 13 investigators met on two occasions to develop guidelines for standardizing the procedures used to test gastrointestinal muscle tone and sensory thresholds using a barostat. General recommendations were: (1) Use a thin-walled plastic bag that is infinitely compliant until its capacity is reached. Maximum diameter of the bag should be much greater than the maximum diameter of the viscus. (2) The pump should be able to inflate the bag at up to 40 ml/sec. (3) Pressure should be monitored inside the bag, not in the pump or inflation line. (4) Subjects should be positioned so that the bag is close to the uppermost surface of the body. (5) For rectal tests, bowel cleansing should be limited to a tap water enema to minimize rectal irritation. Oral colonic lavage is recommended for studies of the proximal colon, and magnesium citrate enemas for the descending colon and sigmoid. (6) If sedation is required for colonic probe placement, allow at least one hour for drug washout and clearance of insufflated air. Ten to 20 min of adaptation before testing is adequate if no air or drugs were used. (7) The volumes reported must be corrected for the compressibility of gas and the compliance of the pump, which is greater for bellows pumps than for piston pumps. (8) Subjects should be tested in the fasted state. For evaluation of muscle tone: (9) The volume of the bag should be monitored for at least 15 min. For evaluation of sensory thresholds; (10) It is recommended that phasic distensions be > or = 60 sec long and that they be separated by > or = 60 sec. (11) Sensory thresholds should be reported as bag pressure rather than (or in addition to) bag volume because pressure is less vulnerable to measurement error. (12) Tests for sensory threshold should minimize psychological influences on perception by making the amount of each distension unpredictable to the subject. (13) Pain or other sensations should be reported on a graduated scale; not "yes-no." The working team recommends verbal descriptor scales, containing approximately seven steps, or visual analog scales in which subjects place a mark on a straight line marked "none" on one end and "maximum" on the other end. (14) It is recommended that subjects should be asked to rate the unpleasantness of distensions separately from their intensity.