The orbitofrontal cortex is part of the prefrontal cortex, situated underneath it, just above the eyes. Many sensory projections lead to it - it has taste input from primary taste cortex, olfactory input, visual input from the inferior temporal cortex, auditory input, and somatosenory (touch etc.) input.

Many studies on the orbitofrontal cortex have been conducted on monkeys due to the similarity of the monkey brain to the human brain, and to circumvent the obvious ethical issues of conducting invasive experiments on humans. By lesioning the orbitofrontal cortex (OFC) in macaques, an insight into its function has been obtained. For instance, lesioned animals are impaired at tasks which involve learning which stimuli are rewarding and which are not. Furthermore, it has been found that they cannot alter their behaviour towards objects when reinforcement contingencies are changed.

Passingham (1975) found that monkeys with lateral OFC lesions were impaired in Delayed Match to Sample (DMS) and Delayed Non-Match to Sample (DNMS) tasks. In these tasks, a monkey sees a stimulus (e.g. a square) on a screen, then that stimulus is taken away. After a short delay, another stimulus is presented. In the DMS task, if the second stimulus matches the first, then the monkey must make a response in order to get a reward. In the DMNS task, rewards are given if the monkey makes a response when the second stimulus is different to the first. Evidence suggests that the OFC plays a role in forming or using stimulus-reward associations.

Monkeys with OFC damage undergo emotional changes, such as decreased aggression to humans and can be sexually indiscriminate. Such behaviour might be explained in terms of impaired stimulus-reward associations. Similar effects are seen in humans who have suffered OFC damage, perhaps from a stroke or an accident of some kind. Humans with OFC damage tend to be euphoric, irresponsible, and have a lack of affect, among many other problems.

The location of the OFC does make it a prime location to form stimulus-reward associations. Rolls (1989) found that taste input is a major representation in the OFC, and furthermore, taste, olfactory and somatosensory stimuli can act as primary reinforcers, meaning that they can act to reward or punish without any learning about the nature of the stimulus having to take place. Such things can be readily associated with inputs from the ventral visual processing stream and auditory cortex, providing so-called "what" information to the OFC. This allows the OFC to associate object information with reward or punishment information (Rolls et al., 1996; Thorpe et al., 1983). A good example of this in humans is that it has been found that face information is represented, which allows for complex associations such as using faces as social reinforcement.

A very important function of the OFC is that it can learn associations in as little as one trial. Furthermore, if reward and punishment associations are reversed, these can be learnt in one trial. This is important for emotional responses, as these are often used so that changes in behaviour towards objects and people (or other animals) can be made immediately depending on, and to adapt to, the current situation. In humans with OFC damage, this stimulus-reward reversibility has been demonstrated as such patients are unable to succeed at the Wisconsin Card Sort Test, as they cannot adapt their sorting strategy as the sorting rule is changed by the experimenter. Without any positive feedback, the patient cannot learn.

The amygdala is another area that has very similar inputs to the OFC, and also forms stimulus-reward associations. However, it is not very good at changing these associations the way the OFC can. The amygdala is good at forming a fundamental association though (e.g., fear responses in relation to electric shocks).