Conjugates of horseradish peroxidase with myelin basic protein (BP) of guinea pig or Lewis rat were used to identify antibody-containing cells in draining lymph nodes during experimental allergic encephalomyelitis (EAE). Peroxidase activity was revealed for light and electron microscopic preparations with the diaminobenzidine reaction of Graham and Karnovsky. Basic proteins (BP) were also iodinated with (125)I for determination of circulating antibody against BP by radio-immunoassay of (125)I BP using coprecipitation with antirat IgG or with antirat serum proteins. Encephalitogenicity was lost after conjugation of guinea pig BP or Lewis rat BP with peroxidase, whereas iodination did not affect the encephalitogenicity of guinea pig or Lewis rat BPs. EAE was induced in Lewis rats with guinea pig or Lewis rat spinal cord BPs in complete Freund's adjuvant. Draining lymph nodes were studied by light and electron microscopy during the course of the immune reaction, and cells with specific antibody against BP were identified with the use of BP-horseradish peroxidase conjugates. Lymph node sections from animals immunized with high antigen doses (500 mug) showed numerous plasma cells with intracellular antibody against BP in medullary cords 10 days after immunization and 4 days prior to histologic appearance of EAE. Numbers of positive cells correlated with levels of circulating antibody against BP. Immunization with a low antigen dose (5 mug) resulted in EAE, few or no antibody-containing cells, and significantly lower levels of circulating antibody. Brown Norwegian rats, a strain resistant to EAE, immunized with 500 mug of BP had positive cells in draining lymph nodes and high levels of circulating antibody against BP in the absence of histologic evidence of EAE. Lewis rats injected with Lewis rat small BP failed to develop EAE. Nevertheless, these animals showed levels of circulating antibody and antibody-containing cells similar to those of animals which developed EAE after injection of the mixture of Lewis rat large and small BP. It is concluded that although the BP-peroxidase labeling method reveals cells with specific anti-BP antibody, these cells are probably unrelated to EAE. The lack of correlation between EAE induced by low antigen doses and levels of circulating anti-BP antibody (determined with the use of highly encephalitogenic (125)I-BP) suggests that effector cells can be stimulated at low antigen doses, but higher antigen doses are required to induce the production of levels of circulating antibody detectable by the method of immune coprecipitation.