Background: Sacral nerve stimulation (SNS) is emerging as a potential treatment for patients with constipation. Although SNS can elicit an increase in colonic propagating sequences (PSs), the optimal stimulus parameters for this response remain unknown. This study evaluated the colonic motor response to subsensory and suprasensory SNS in patients with slow-Transit constipation. Methods: Patients with confirmed slow-Transit constipation were studied. Either a water-perfused manometry catheter or a high-resolution fibre-opticmanometry catheter was positioned colonoscopically to the caecum. A temporary electrode was implanted transcutaneously in the S3 sacral nerve foramen. In the fasted state, three conditions were evaluated in a double-blind randomized fashion: sham, subsensory and suprasensory stimulation. Each 2-h treatment period was preceded by a 2-h basal period. The delta (Δ) value was calculated as the frequency of the event during stimulation minus that during the basal period. Results: Nine patients had readings taken with a water-perfused catheter and six with a fibre-optic catheter. Compared with sham stimulation, suprasensory stimulation caused a significant increase in the frequency of PSs (mean(s.d.)Δvalue-1·1(7·2) versus 6·1(4·0) PSs per 2 h; P = 0·004).Nomotor response was recorded in response to subsensory stimulation compared with sham stimulation. Compared with subsensory stimulation, stimulation at suprasensory levels caused a significant increase in the frequency of PSs (P = 0·006). Conclusion: In patients with slow-Transit constipation, suprasensory SNS increased the frequency of colonic PSs, whereas subsensory SNS stimulation did not. This has implications for the design of therapeutic trials and the clinical application of the device.