Two studies are reported which investigated the relationship between expectancy of the UCS (shock) and avoidance conditioning of the skin conductance response (SCR). Subjects were allowed to avoid the shock UCS if they gave criterion magnitude SCRs to a warning tone preceding the shock. Expectancy of the UCS was monitored on a trial-by-trial basis during the avoidance conditioning process. It was found in Experiment I that SCR conditioning occurred only among subjects who had accurate expectancies of shock omission relative to their yoked controls. However, those subjects who were able to accurately predict shock omission were not able to verbalize the response-reinforcement contingency during a postconditioning questionnaire. Experiment II revealed that only subjects with accurate expectancies concerning the shock behaved differently from their yoked controls when given informative instructions about the response-reinforcement contingency. These data indicate that individual differences in the accuracy of UCS expectancy are positively related to SCR avoidance conditioning. This relationship has gone undetected in previous research because of the exclusive reliance upon verbal postconditioning questionnaires.