The authors determined the duration of analgesia, toxicity, and neuraxial distribution of liposomal morphine after intrathecal administration in the mouse. Analgesic duration was determined using the tail-flick test after intrathecal injection of 12.5, 25, or 50 micrograms of plain or liposomal morphine (n = 6 mice/dose/formulation). Toxicity of the formulations was compared by estimating LD50. Neuraxial morphine distribution was determined after 20 micrograms of plain or liposomal morphine. The excised spinal cord and brain were divided into five segments at 1 min, and at 1, 4, and 8 h after injection for both formulations. In addition, for the liposomal morphine, similar sections were obtained at 24 h (n = 6 mice/formulation/time point). Segmental morphine concentration was quantified using radioimmunoassay. Liposomal encapsulation significantly prolonged duration of analgesia for the 25-micrograms (13.4 +/- 1.64 [SE] vs 4.1 +/- 0.5 h) and 50-micrograms doses (16.8 +/- 4.0 vs 4.6 +/- 1.0 h). The estimated LD50 was 200 (confidence interval 151-257 micrograms) for plain morphine, but was not determinable for the liposomal formulation, since no deaths occurred at the largest dose level which could be tested (371 micrograms). For plain morphine, the drug was not confined to a specific neuraxial segment, and segmental levels declined rapidly. After liposomal morphine, the most morphine was concentrated and persisted in the low spinal cord segment at each time interval. These results show that a single dose of liposomal morphine produces prolonged analgesia with decreased toxicity compared to the plain formulation.