Colonic myoelectric activity was determined using a mucosal clip electrode in normal human subjects, in the basal state and following the administration of prostigmin, pentagastrin, cholecystokinin, or secretin. Slow-wave activity was present 49 +/- 3.0% of the recording time and occurred at two frequencies, 6.5 +/- 0.1 cycles/min and 3.5 +/- 0.1 cycles/min, in the rectum. Spike potentials were infrequent in the basal state. Prostigmin, pentagastrin, or cholecystokinin did not affect the slow-wave activity or frequency, but these agents significantly increased the number of spike potentials (P less than 0.05). Secretin had no effect on myoelectric activity. The electrode system was tested in the cat and showed excellent correlation of the myoelectric activity between the mucosal clip electrode and a serosal electrode. These studies indicate: (1) a mucosal clip electrode provides an accurate method of recording colonic myoelectric activity when compared to a serosal electrode in the cat; (2) in man, slow-wave activity is present only intermittently and exists at two frequencies, approximately 6 cycles/min and 3 cycles/min; (3) gastrointestinal hormones and prostigmin increase spike potential discharge without altering slow-wave activity or frequency.