Inflammatory components play an important part in many diseases of the central nervous system (CNS). Recent evidence suggests that this may also be true of diseases which were previously considered as purely neuro-degenerative. However, it is also clear that inflammatory responses in the CNS differ in many ways from responses in non-CNS tissues. Some of these differences have been demonstrated by the use of animal models. For example, when bacteria are injected into the brain parenchyma, they induce a typical acute inflammatory response. However, unlike in other tissues, bacteria which are not cleared from the brain parenchyma remain undetected by the immune system. Some bacteria, such as bacillus Calmette-Guérin, can persist in the brain parenchyma for months sequestered in microglia and perivascular macrophages. When an animal with an intraparenchymal bacteria deposit is later sensitised peripherally, an immune response is evoked at the site of the deposits. The lesions induced in the CNS parenchyma are T-cell mediated and show characteristics typical of a delayed-type hypersensitivity response. The lesions produce a breakdown of the blood-brain barrier and demyelination. These immune responses are similar to those described for multiple sclerosis lesions. The responses to bacteria are unique to the brain parenchyma. Pathogens injected into the ventricles induce inflammatory responses similar to those in other non-CNS tissues: there is an acute inflammatory response which develops spontaneously into an immune mediated response within the first week.