Recent neurophysiological and pathological studies have led to a reclassification of the diseases that underlie Guillain-Barré syndrome (GBS) into acute inflammatory demyelinating polyradiculoneuropathy (AIDP), acute motor and sensory axonal neuropathy (AMSAN) and acute motor axonal neuropathy (AMAN). The Fisher syndrome of ophthalmoplegia, ataxia and areflexia is the most striking of several related conditions. Significant antecedent events include Campylobacter jejuni (4-66%), cytomegalovirus (5-15%), Epstein-Barr virus (2-10%), and Mycoplasma pneumoniae (1-5%) infections. These infections are not uniquely associated with any clinical subtype but severe axonal degeneration is more common following C. jejuni and severe sensory impairment following cytomegalovirus. Strong evidence supports an important role for antibodies to gangliosides in pathogenesis. In particular antibodies to ganglioside GM1 are present in 14-50% of patients with GBS, and are more common in cases with severe axonal degeneration associated with any subtype. Antibodies to ganglioside GQ1b are very closely associated with Fisher syndrome, its formes frustes and related syndromes. Ganglioside-like epitopes exist in the bacterial wall of C. jejuni. Infection by this and other organisms triggers an antibody response in patients with GBS but not in those with uncomplicated enteritis. The development of GBS is likely to be a consequence of special properties of the infecting organism, since some strains such as Penner 0:19 and 0:41 are particularly associated with GBS but not with enteritis. It is also likely to be a consequence of the immunogenetic background of the patient since few patients develop GBS after infection even with one of these strains. Attempts to match the subtypes of GBS to the fine specificity of anti-ganglioside antibodies and to functional effects in experimental models continue but have not yet fully explained the pathogenesis. T cells are also involved in the pathogenesis of most or perhaps all forms of GBS. T cell responses to any of three myelin proteins, P2, PO and PMP22, are sufficient to induce experimental autoimmune neuritis. Activated T cells are present in the circulation in the acute stage, up-regulate matrix metalloproteinases, cross the blood-nerve barrier and encounter their cognate antigens. Identification of the specificity of these T cell responses is still at a preliminary stage. The invasion of intact myelin sheaths by activated macrophages is difficult to explain according to a purely T cell mediated mechanism. The different patterns of GBS are probably due to the diverse interplay between antibodies and T cells of differing specificities.