Both the learning of an aversive CS-US pairing in classical fear conditioning and the extinction of a fear response critically rely on the amygdala to occur. Additionally, several lines of evidence suggest that extinction learning is not simply an erasure of the original fear memory, but a distinct and parallel form of learning that suppresses the expression of the conditioned fear response. Thus there appear to be two forms of learning that occur in the basolateral amygdala which have opposing effects on the expression of conditioned fear through the output of the amygdala (the central nucleus). We hypothesize that these two forms of learning will produce differential patterns of second-messenger and transcription factor activation within the sub-populations of excitatory and inhibitory neurons that constitute the associative learning circuit in the basolateral complex of the amygdala (BLA). Furthermore, we hypothesize that by using pharmacologic and genetic manipulations targeting one particular sub-population of BLA neurons, we may be able to selectively alter one of these two types of learning while leaving the other intact.