The effects of naloxone and diprenorphine, opiate antagonists with different receptor-binding properties, and the putative kappa-receptor agonists, ketocyclazocine and ethylketocyclazocine (EKC), were studied on food-reinforced responding in rats. Behavior was maintained under a multiple-component 1-min variable-interval schedule in which 12-min periods of unpunished responding alternated with 4-min periods in which each response was punished by a brief electric footshock. Daily sessions were 1 h. Naloxone (0.01-10 mg/kg) decreased unpunished responding only slightly; punished responding was decreased significantly to 66% of control by 10 mg/kg. Diprenorphine (0.01-10 mg/kg) did not affect unpunished responding and increased punished responding dose-dependently to as much as 190% of control. EKC (0.01-1.0 mg/kg) decreased unpunished and punished responding dose-dependently and comparably, whereas ketocyclazocine (0.01-1.0 mg/kg) decreased unpunished responding but did not significantly affect punished responding. Diprenorphine was more potent than naloxone in blocking the decreases in responding produced by the kappa agonists. Differences in the behavioral effects of naloxone and diprenorphine appear to reflect the different receptor-binding properties of the two opiate antagonists.