Previous attempts at lie detection, such as the polygraph, have relied on physiological arousal to identify deception--but these responses have not proven to be as reliable as is necessary for widespread use. Conditioning procedures have been shown to increase the discriminative physiological arousal exhibited during deception, but have targeted only instances of deception where the experimenter knows the subject is being deceptive, and the subject knows the experimenter knows (public instances of deception). To be practically useful, such a method must improve the responses subjects exhibit when "only" they know they are being deceptive (private deception). The present experiment used conditioning procedures to increase skin conductance to instances of public deception, and then tested whether these responses generalized to deceptive answers regarding a mock crime when only the subjects knew they were being deceptive. The results indicated that these responses can generalize across the public/private barrier; participants responded more strongly following instances of deception that only they were aware of. (Contains 4 figures.)