The relation between Guillain-Barré syndrome and chronic inflammatory demyelinating polyneuropathy is discussed. Most likely they represent parts of a continuum, arbitrarily separated by their time course. Within the concept of chronic inflammatory demyelinating polyneuropathy the presence of a monoclonal gammopathy of undetermined significance is discussed. The pathogenesis of inflammatory demyelinating polyneuropathies has not been elucidated yet, but involvement of the immune system has been firmly established. Preceding infections, especially with Campylobacter jejuni, and the analysis of antiganglioside antibodies lend new support to the hypothesis of molecular mimicry between epitopes on infectious agents and peripheral nerve constituents as one of the mechanisms in Guillain-Barré syndrome. In the future, a further classification of individual patients based on clinical, epidemiological, electrophysiological, pathological, microbiological, and immunological criteria may give a basis for more individualized treatment strategies. In Guillain-Barré syndrome the efficacy of high-dose intravenous immune globulin treatment was established after earlier positive findings with plasma exchange; immune globulins are easier to administer and may be superior. Even with these treatments it should be anticipated that one fourth of patients after immune globulin treatment and one third of patients after plasma exchange will show further deterioration in the first 2 weeks after onset of treatment. Despite this, just one treatment course usually is indicated in the individual patient, and no valid arguments were found to switch to the other treatment modality. In chronic inflammatory demyelinating polyneuropathy, prednisone, plasma exchange, and immune globulins are effective in a proportion of patients. The last two are equally effective. Patients may respond to one of these if a previous treatment failed, and here switching therapy may be effective due to the chronic course of the disease. Complexity and costs make plasma exchange the last choice. Whether prednisone or immune globulin is the first choice depends on the speed of recovery and the estimation of long-term loss of quality of life due to side effects of prednisone versus the costs of immune globulins. The mechanism of immune globulins in inflammatory polyneuropathies is discussed. There is evidence that idiotypic-antiidiotypic interaction may play a role, but several other mechanisms also may be involved.