Tramadol, a mixed mu-opioid agonist and a monoamine-reuptake blocking analgesic, has been supposed to have little effect on propulsive gastrointestinal motility. However, this has not been specifically studied in man. Following institutional approval, 18 human volunteers were given 50 mg of tramadol, tilidine/naloxone, and codeine, respectively, in a double-blind randomised cross-over design. Additionally, 12 further volunteers were given 100 mg of each opioid in a double-blind, randomised fashion, followed by measurement of gastrocoecal transit time. Gastrointestinal transit time was measured using the lactulose H(2)-breath test. A threefold increase in end-expiratory hydrogen when compared to the control value was considered the end point of gastrocoecal transit. At the low dose (50 mg) the three opioids did not differ significantly with regard to their effect on gastrointestinal motility. Gastrocoecal transit time was 90.8 (+/- 10.1 SEM) min for tramadol, 100.6 (+/- 9.8 SEM) min for tilidine/naloxone, and 104.2 (+/- 8.7 SEM) min for codeine. Doubling the dose of each opioid resulted in an increase in mean gastrocoecal transit, namely 97.8 (+/- 11.2 SEM) min for tramadol, 129.2 (+/- 12.2 SEM) min for tilidine/naloxone and 135.9 (+/- 9.2 SEM) min for codeine. The increase in gastrocoecal transit time was significant (P < 0.01) for high doses of tilidine/naloxone and codeine in contrast to the effect of the low doses. This lesser constipation effect may be due to the reduced affinity of tramadol to the mu-opioid receptor. Sedation was significantly higher for codeine after 50 mg (P < 0.05) and 100 mg (P < 0.005) than for tilidine/naloxone and tramadol. Vertigo was significantly higher after 50 mg (P < 0.05) and 100 mg (P < 0.005) of tilidine/naloxone and codeine than after tramadol. Perspiration was significantly higher after tramadol 100 mg (P < 0.005) than after tilidine/naloxone and codeine. Sedation is considered a typical symptom of analgesics interacting with centrally located opioid receptor sites. The higher incidence of perspiration after tramadol suggests that monominergic pathways may be involved in thermoregulation. In conclusion, the opioids tilidine/naloxone and codeine at the doses used significantly prolong gastrointestinal transit time in the high-dose range. Since tramadol does not induce a dose-related increase in gastrocoecal transit time, it may be a useful analgesic in patients who are prone to developing constipation during high-dose opioid therapy.