The effect of liposome encapsulation on the analgesia produced by intrathecally administered alfentanil was examined in the rat. In rats prepared with chronic intrathecal catheters, alfentanil in doses of 1-50 micrograms was administered intrathecally in either saline or in multilamellar liposomes (dipalmitoylphosphatidylcholine and cholesterol). Animals were then tested for analgesia by hot-plate and paw-pressure tests. A second group of animals received intrathecal injections of 30 micrograms alfentanil in saline or liposomes, and blood samples were obtained at 5, 15, 45, and 135 min thereafter for measurement of alfentanil plasma concentrations. The liposome preparation markedly prolonged spinal analgesia in the paw-pressure test and to a lesser extent in the hot-plate test. Neither the time to peak analgesia nor the intensity of analgesia differed between the saline and liposome groups. Liposome encapsulation significantly reduced the peak alfentanil plasma concentration at 5 min and prolonged the period in which low but measurable levels of alfentanil could be measured in plasma. These pharmacokinetic data demonstrate that liposome encapsulation resulted in a slow but prolonged appearance of free alfentanil into a diffusible pool available for uptake into the spinal cord. Consistent with the lower peak plasma concentration of alfentanil, the liposome group demonstrated a significantly lower incidence of catalepsy, indicating less systemic redistribution of alfentanil to supraspinal sites. Liposome encapsulation thus appears to produce a significant reduction in peak plasma concentration with a concomitant reduction in systemic side effects and an increase in the duration of action for a given intrathecal dose of the otherwise rapidly cleared alfentanil.