1. We investigated the involvement of muscle tone and circular muscle (CM) contraction in peristalsis in isolated guinea-pig small intestine. 2. A segment of jejunum (∼13 cm) was mounted into a three chambered partitioned bath. Peristaltic waves were initiated in the oral chamber either by: (1) infusing fluid into the oral end of the jejunum; the ejected fluid was diverted via a cannula from reaching the intermediate and anal chambers, or by (2) intraluminal balloon distension of the empty oral segment. Tension of the circular muscle was measured in all three chambers. 3. Peristaltic waves elicited by fluid infusion were evoked at an abrupt threshold. 4. In contrast, peristaltic waves elicited by distension could be graded in amplitude according to stimulus intensity. 5. Peristaltic waves evoked in an empty intestine exhibited similar propagation velocities to peristaltic waves associated with fluid propulsion. 6. Nifedipine (200-400 nM) added to the intermediate chamber to block muscle contraction did not prevent peristaltic waves elicited by either stimulus from propagating into the anal chamber, although their amplitude was attenuated. Nifedipine to the site of stimulation (oral chamber) abolished peristaltic waves generated by either stimulus. 7. Tetrodotoxin (1-2 μm), or a low Ca2+-high Mg2+ solution to the intermediate chamber abolished the propagation of peristalsis from the oral to anal chambers. 8. In conclusion, graded peristaltic waves can occur in an empty intestine. Therefore peristalsis is not necessarily an 'all-or-none' phenomenon. Peristalsis depends on the spread of nervous activity along the bowel, rather than the reactivation of neural circuits caused by displacement of fluid in the lumen. However, local muscle tone and contraction are important for the initiation and maintenance of peristaltic propagation.