In rats, intrathecal alfentanil, lofentanil, sufentanil, fentanyl, and morphine produced dose-dependent elevations in the hot-plate and tail-flick latencies and a powerful suppression of the writhing response. The slopes of the monotonic dose-response curves for the Five opioids did not differ significantly. In terms of the hot-plate ED50after intrathecal injection, the order of potency was as follows: lofentanil (210), sufentanil (29), fentanyl (3), morphine (1), and alfentanil (1). Comparable results were observed in the tail flick. The duration of action was proportional to dose. However, at doses that produced an equal magnitude of inhibition, the duration of action was lofentanil > morphine > sufentanil > alfentanil ≥ fentanyl. Systemically administered naloxone (0.03–1 mg/kg, sc) resulted in dose-dependent antagonism of the antinociceptive effect of intrathecal morphine, fentanyl, alfentanil, and sufentanil. In contrast, intrathecal lofentanil was extremely resistant to antagonism by naloxone. In cats, similar dose-dependent blockade of the thermally evoked skin-twitch response was observed after intrathecal morphine, sufentanil, alfentanil, and fentanyl. As in the rat, the slope of the monotonic dose-response curves did not differ. The relative potency and duration of action after equipotent intrathecal doses were similar to those observed in the rodent. These results suggest that sufentanil, alfentanil, and fentanyl exert their analgesic effectsin vivoat a spinal cord site that has properties comparable to those of the site acted upon by morphine. Except for catalepsy in rats, no major behavioral dysfunctions were noted at the ED50dose of any of the drugs administered. No abnormal morphologic effects of acutely or chronically administered alfentanil and sufentanil were seen, aside from an inflammatory reaction secondary to catheter placement.