Pigeons acquired a different four-response chain each session by responding sequentially on three keys in the presence of four colors. The response chain was maintained by food presentation under a fixed-ratio schedule. Errors produced a brief timeout but did not reset the chain. When either morphine or naloxone was administered alone, the overall response rate decreased with increasing doses. The rate-decreasing effect was accompanied by an increase in percent errors with morphine but not with naloxone. Both effects of morphine were antagonized by doses of naloxone that were ineffective when given alone. The antagonism was selective in that naloxone (3 mg/kg) completely blocked the error-increasing effect but not the rate-decreasing effect of the higher doses of morphine. The view that naloxone is a specific narcotic antagonist was supported by the finding that naloxone failed to antagonize the rate-decreasing and error-increasing effects of d-amphetamine, pentobarbital, and phencyclidine.Key words: repeated acquisition, response chains, fixed-ratio schedule, drug antagonism, morphine, naloxone, key peck, pigeonsThe effects of drug combinations on complex operant behavior have received little experimental attention. In one of the few studies in this area, Thompson (1980) examined the joint effects of d-ainphetamine and chlorpromazine on behavior in a repeated-acquisition task. Pigeons acquired a different fourresponse chain each session by responding sequentially on three keys in the presence of four colors. The response chain was maintained by food presentation under a fixed-ratio (FR) schedule. When d-amphetamine was administered alone, the overall response rate decreased and the percent errors increased with increasing doses. When a small dose of chlorpromazine, which was ineffective when given alone, was administered in combination with d-amphetamine, the rate-decreasing effect was antagonized. The antagonism was selective, however, in that the error-increasing effect of d-amphetamine was augmented by chlorpromazine. The nature of the joint effect of the two drugs thus depended on the behavioral measure: rate vs. accuracy (cf. Branch, 1974). The present research examined the possibility that morphine and naloxone would interact in a manner dissimilar to d-amphetamine and chlorpromazine in a repeated-acquisition task (Thompson, 1980). Morphine and naloxone were chosen for study because the effects of separate and combined administration of these two drugs on schedule-controlled performance have been investigated extensively in pigeons (see review by McMillan, 1974). For example, when the responding of pigeons is maintained under an FR schedule of food presentation, it has been found that 1) increasing the dose of either morphine or naloxone alone decreases the overall response rate and 2) the rate-decreasing effect of morphine can be antagonized by doses of naloxone that have no effect when given alone (Downs