Two main hypotheses support the contention that type I diabetes is an autoimmune disease, namely "the cytokine hypothesis" and "the T-lymphocyte hypothesis". Various strategies can be used at diagnosis with a view to stopping beta-cell destruction or at least attenuating the process. This review discusses the use of antioxidants based on the idea that free oxygen radicals are important mediators. Experimental animal models have indicated that several antioxidants may prevent diabetes, although in humans only nicotinamide has been shown to have some effect on preventing the disease in high-risk individuals and to produce a slight effect on residual insulin secretion in newly diagnosed patients. Bearing in mind these considerations, we tried a cocktail of several antioxidants at high dosage. As the code of this randomized double-blind study is not broken, results cannot be given, but preliminary observation indicates that there has been no dramatic increase of complete remission. Based on the hypothesis that type I diabetes is a T-lymphocyte-mediated disease, lymphocyte photopheresis may be useful. Photopheresis, comprising the treatment of lymphocytes by a combination of the light-activated drug methoxypsoralen and UVA irradiation, has been shown to be effective in the treatment of some other autoimmune disease. One hypothesis regarding its efficacy holds that the method causes changes in the antigenicity of the treated lymphocyte clones which cause a vaccination-like effect when these cell lines are retransfused at repeated intervals into the patient. Nothing to date is known about its effect in diabetes, although a double-blind randomized placebo-controlled study has been commenced.