Ventilatory sensitivity to CO2 was measured at various times (0.5, 1, 3, 6, 10, 16, and 22 h) in 10 healthy young volunteers after 10 mg of morphine sulfate in 10 ml of saline injected intravenously (IVm) or by the epidural route (Em). The two randomized study sequences were completed 2-4 weeks apart. Ventilatory variables studied were resting end-tidal CO2 (PetCO2) measured before each rebreathing maneuver; slopes of the ventilatory response curve (sVE) and position of the curve, calculated as the ventilation sustained for a fixed stimulus of PetCO2 = 54 torr (VE54). Additionally, linear regressions were calculated for tidal volumes (VT) and respiratory rates (RR) during the rebreathing test, yielding sVT, VT54, sRR, and RR54. CO2-response curves were maximally depressed following IVm at the 0.5-h study period, while after Em, maximal respiratory depression was at the 6- and 10-h study period. Significantly greater depression after Em was demonstrated between 3 and 22 h by one or more of the following parameters: PetCO2, sVE, VE54, VT54, and sRR. The results indicate substantial differences in magnitude, duration, and characteristics of the depression of the CO2 chemosensitivity between the two modes of administration of morphine, quite separate from the differences observed for serum morphine levels in these volunteers.