The acquisition of a conditioned response to a stimulus when it is paired with a reinforcer is retarded if the stimulus has previously been repeatedly pre-exposed in the absence of the reinforcer. This effect, called latent inhibition, has previously been found to be insensitive to lesions of the medial prefrontal cortex (mPFC) in rats. Using an on-baseline conditioned emotional response procedure, which is especially sensitive to small variations in the absolute magnitude of latent inhibition, we found increased latent inhibition following excitotoxic lesions of the mPFC (Experiment 1) or the ventral mPFC alone (Experiment 2) as compared with sham-operated control rats. Lesions restricted to the dorsal mPFC, however, were without effect (Experiment 2). These results are consistent with those of experiments employing another type of interference procedure, extinction. Together, these findings suggest that when different contingencies between a stimulus and a reinforcer are established in separate learning phases, lesions to the ventral mPFC result in increased interference between first-learned and second-learned contingencies. As a consequence, retrieval of the second-learned contingency is impaired, and performance is dominated by the first-learned contingency. These findings are discussed in light of the use of latent inhibition to model cognitive deficits in schizophrenia.