We studied the effects of the immunosuppressant sodium fusidate (fusidin) on murine immunoinflammatory diabetes mellitus (DM) induced by multiple low doses of streptozotocin (SZ). Fusidin was given by gavage to three strains of mice (C57KsJ, C57BL/6, CD1) at doses 10 or 100mg/kg body weight every other day. The drug was administered as an early or late prophylactic regime starting either 1 day prior to the first or after the fifth and last injection of SZ. In both situations the largest dose of fusidin successfully reduced the clinical, chemical and histological signs of DM, the treated mice having significantly lower glycaemic values and milder (often absent) insulitis compared with sham-treated animals or controls given SZ alone. The antidiabetogenic effect was long-lasting as it was maintained up to 1 month after cessation of therapy. In contrast, fusidin prophylaxis failed to prevent development of hyperglycaemia acutely induced by one single and high (160mg/kg) dose of SZ, which is a model of DM primarily due to the toxic action of SZ on the β cells and does not involve immunopathogenetic mechanisms. On day 14 after SZ, fusidin markedly altered the circulating cytokine profile inducedin vivo by ConA, reducing the levels of IFN-γ, IL-2 and TNF-α and augmenting the level of IL-6. However, only the inhibitory effect of the drug on the synthesis/release of IFN-γ seemed to be causally related to its capacity to counteract the SZ-induced DM. In fact, the disease was prevented by a neutralizing monoclonal antibody (mAb) against IFN-γ, but not by anti-IL-2 receptor mAb, a soluble form of TNF-receptor type 1 or recombinant human IL-6. The prevention of disease by fusidin was also partly reversed by exogenously administered recombinant mouse IFN-γ. The data provide furtherin-vivo evidence for the anti-diabetogenic and immunomodulatory properties of fusidin and indicate that this drug could have a role in prevention and treatment of human type 1 DM.