Rats of both sexes from the genetically selected Roman High Avoidance (RHA), the Roman Low Avoidance (RLA) and the Roman Control (RCA) strains were given one 30-min session of two-way escape-avoidance conditioning (30 trials) in a shuttle box with a buzzer as the conditioned stimulus and shock as the unconditioned stimulus in a factorial design involving three time intervals (0, 15 and 30 min) following one subcutaneous injection of saline or of nicotine in five doses (0.05, 0.1, 0.2, 0.4, or 0.8 mg/kg of body weight). Six measures relating to performance were analysed in whole or part. While the number of avoidance responses showed the expected strain differences, no effect of dose or delay interval could be detected. Avoidance and escape latencies and intertrial activity showed some effects of these variables, especially in interaction with sex and strain. Dose determined pre-sessional activity, but its effect was strain dependent. The failure to confirm previous findings is discussed in the context of other instances in the literature, and the results of an operant experiment confirming the continuing sensitivity of the Roman strains to the effects of nicotine are reported.