In the therapeutic manoeuvre termed "lymphocyte vaccination", activated lymphocytes capable of transferring an autoimmune disease are instead attenuated and given in vaccine form. We have previously shown that such a therapy administered to non-obese diabetic (NOD) mice at 6 weeks of age prevents diabetes mellitus. To assess whether this therapy has potential clinical relevance, in the present study lymphocyte vaccination was applied in NOD mice in 3 weekly doses commencing in the immediate prediabetic period (age 12 weeks), when insulitis is advanced and diabetes incipient. Of 30 NOD mice receiving active vaccine (composed of attenuated lymphocytes from diabetic NOD mice) 13 (43.3%) remained non-diabetic to the age of 30 weeks, in comparison with 2 of 30 (6.7%; p < 0.01) mice receiving a control vaccine (composed of attenuated lymphocytes from non-diabetic NOD/B10 mice) and 5 of 26 (19.2%; p < 0.01) mice receiving saline carrier alone. Moreover, in an additional group of 10 NOD mice receiving active vaccine weekly between 12 and 30 weeks, 8 remained diabetes free at the end of the treatment. The most notable effect of the vaccine was that the delay in diabetes onset was accompanied by a reduction in insulitis and in some cases a complete absence of infiltrating lymphocytes at 30 weeks of age. Immunocytochemistry indicated that when present, islet infiltrating lymphocytes in non-diabetic mice that received active vaccine showed significantly reduced staining for interferon-gamma, compared with the infiltrate seen in diabetic mice receiving the control vaccine or saline. This study demonstrates that the rapid progression to diabetes typically seen in 12-week-old NOD mice can be delayed by lymphocyte vaccination, supporting the possibility that a vaccine composed of attenuated autologous peripheral blood lymphocytes could be effective in high risk first degree relatives of patients with insulin dependent diabetes mellitus.