Second-order conditioning of social approach to a female conspecific in male Japanese quail was investigated in four experiments. Subjects that received paired first-and second-order trials acquired second-order conditioning in both Experiments 1 and 2. In contrast, subjects that received paired first-order but unpaired second-order trials, and subjects that received unpaired first-order but paired second-order trials, did not acquire second-order conditioning. In Experiment 3, subjects for whom the first-order conditioned stimulus was presented in extinction showed second-order conditioning comparable to that shown by subjects in a control group that did not receive the extinction procedure. In Experiment 4, subjects approached a second-order stimulus less when sexually satiated than when sexually deprived. These flndings suggest that second-order sexual conditioning in quail is mediated by an association of the second-order stimulus with a representation of the unconditioned stimulus.Second-order conditioning has been ofconsiderable historical interest because it putatively increases the range of situations in which Pavlovian conditioning may occur. First-order Pavlovian conditioning requires the pairing of a conditioned stimulus (CS) with an unconditioned stimulus (US), such as food, water, or shock, that is biologically important to the subject. This limits first-order conditioning to situations in which subjects encounter unconditioned stimuli. In contrast, in second-order conditioning a previously conditioned stimulus (CS 1) is used in the role of a US to produce conditioning to a new stimulus (CS2). Therefore, second-order conditioning can occur in situations that do not involve exposure to an unconditioned stimulus, provided that the subjects previously receive first-order conditioning.Most modem studies ofsecond-order conditioning have involved fear conditioning in rats (e.g., Rizley & Rescorla, 1972) or appetitive conditioning in rats and pigeons (e.g., Holland & Rescorla, 1975b;Rescorla, 1979). The present experiments were conducted to extend the study ofsecondorder conditioning to sexual behavior.Second-order sexual conditioning was previously demonstrated in rats by Zamble, Hadad, Mitchell, and Cutmore (1985), but they measured conditioning only indirectly by demonstrating shorter ejaculation latencies following exposure to the second-order stimulus. Sexual conditioning has been investigated more extensively in male Japanese quail (see Domjan, 1994, for a review). Studies with Japanese quail have confirmed that, as in rats, sexually conditioned stimuli can facilitate copulatory behavior (e.g., Domjan, Greene, & North, 1989). In addition, these experiments have identified circumstances in which first-order conditioned stimuli can also directly elicit various conditioned responses (e.g., Akins, Domjan, & Gutierrez, 1994).Male quail approach and remain near visible but physically inaccessible females (Domjan & Hall, 1986) and also come to approach and remain near visual stimuli that have been paired w...