During Pavlovian conditioning the expression of a conditioned response typically serves as evidence that an association between a conditioned stimulus (CS) and an unconditioned stimulus (UCS) has been learned. However, learning-related changes in the unconditioned response (UCR) produced by a predictable UCS can also develop. In the present study, we investigated learning-related reductions in the magnitude of the unconditioned skin conductance response (SCR). Healthy volunteers participated in a differential conditioning study in which one tone (CS+) was paired with a loud white-noise UCS and a second tone (CS-) was presented alone. In addition, probe trials that consisted of UCS presentations paired with the CS+ (CS + UCS) and CS- (CS - UCS), as well as presentations of the UCS alone were included to assess UCR diminution. SCR and participants' expectations of UCS presentation were monitored during conditioning. Greater diminution of the UCR was observed to the UCS when it followed the CS+ compared to when it followed the CS- or was presented alone. Further, UCR amplitude showed an inverse relationship with the participants' ratings of UCS expectancy. However, conditioned UCR diminution was also observed independent of differential UCS expectancies. Our findings demonstrate conditioned diminution of the unconditioned SCR. Further, these findings suggest that although UCR amplitude is modified by conscious expectations of the UCS, conditioned diminution of the UCR can be expressed independent of learning-related changes in these expectations.