Using a new quantitative immunoblot technique (MegaBlot and artificial neural network) sera of rats with lens-induced uveitis (LIU) were tested against Western blots (WB) of sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) separations of protein fractions from rat lenses. The blots were scanned using digital image analysis and evaluated by multivariate analysis of discriminance and an artificial neural network. Five groups of LIU rats (male Wistar rats, n = 28) were investigated: no treatment, single doses of methylprednisolone (MPR; 7.5 mg/kg body weight, intravenous), allopurinol (AL; 50 mg/kg body weight, intravenous), a combination of both drugs (AL and MPR), and repeated application of AL (ALFR; 50 mg/kg body weight intravenous every 2 weeks during the immunization period and a daily dose of approximately 25 mg/kg body weight, orally). ALFR, MPR, and MPRAL show statistically significant immunological effects in LIU treatment. Classification of WBs using both discriminance analysis or neural network resulted in 100% correct assignment; and 82% (discriminance analysis) and 86% (neural network) of WBs were correctly assigned using only unknown blots for classification (not studied before). The MegaBlot technique is able to detect therapeutically induced changes in autoantibody repertoires allowing therapeutic control and a prognostic view of treatment.