Multiple sclerosis (MS) can be divided into 4 clinical forms: relapsing-remitting (RR), primary progressive (PP), secondary progressive (SP), and progressive relapsing (PR). Since PP-MS is notably different from the other forms of MS, both clinically and pathologically, the question arises whether PP-MS is immunologically similar to the other forms. The pathogenesis of the PP-MS remains unclear, partly due to a lack of highly relevant animal models. Using an encephalitogenic peptide from myelin oligodendrocyte glycoprotein (MOG)92-106, we have established animal models that mimic different forms of MS in 2 strains of H-2s mice, SJL/J and A.SW. We induced experimental allergic encephalomyelitis (EAE) using MOG92-106 in the presence or absence of supplemental Bordetella pertussis (BP). Although, SJL/J mice developed RR-EAE whether BP was given or not, A.SW mice developed PP-EAE without BP and SP-EAE with BP. Histologically, SJL/J mice developed mild demyelinating disease with T cell infiltration, while A.SW mice developed large areas of plaque-like demyelination with immunoglobulin deposition and neutrophil infiltration, but with minimal T cell infiltration. In A.SW mice without BP, high titer serum anti-MOG antibody was detected and the anti-MOG IgG2a/IgG1 ratio correlated with survival times of mice. We hypothesized that, in A.SW mice, a Th2 response favors production of myelinotoxic antibodies, leading to progressive forms with early death. Our new models indicate that a single encephalitogen could induce either RR-, PP-, or SP- forms of demyelinating disease in hosts with immunologically different humoral immune responses.