The orbitofrontal cortex (OFC) has received a lot of attention in the past decade, and has been implicated in a wide range of functional roles such as response inhibition, encoding predicted value, encoding the sensory properties of expected outcomes, and encoding “latent states” in model based reinforcement learning. These hypotheses are predominantly modelled on deficits in outcome devaluation, and reversal learning following disruption to OFC function, while leaving initial acquisition learning intact. However, all of these key findings have had replication issues. I will present a series of experiments that explores the role of the rodent lateral OFC in (1) outcome devaluation, (2) reversal learning, and (3) simple Pavlovian learning procedures. These experiments reveal the importance of the rodent lateral OFC in a specific associative pathway allowing current values to update Pavlovian actions. Furthermore, a functional dissociation is found between anterior and posterior subregions of the lateral OFC. These findings also provide a possible solution to the often contradictory findings in the OFC literature that involve data taken from a range of species (rodents, non-human primates, and humans), using different experimental protocols, and often involving comparing across different orbital sub regions