The toxic oil syndrome (TOS) occurred in Spain in 1981 as a result of ingestion of oil mixtures containing aniline-denatured rapeseed oil. The disease afflicted almost 20000 people, resulted in more than 400 deaths, and mimicked an autoimmune disease in all patients. Phenilamine-propanediol (PAP) has been implicated as a possible etiologic agent of TOS but absence of an acceptable animal model to evaluate the autoimmune potential of the 'case oil' has hindered identification of the actual etiologic agent(s). The purpose of this study was twofold; (1) to develop an animal model of human disease to investigate the immunological etiology and pathogenesis of TOS and (2) to determine if the 'case oil' responsible for TOS and/or two synthesized oils either induced or exacerbated the systemic autoimmune disease that occurs spontaneously in the MRL/lpr mouse. The oils tested were a denatured rapeseed oil collected from a family (case oil) who were affected by the TOS (CO756), a rapeseed oil denatured with 2% aniline and enriched with a mixture of diesters of PAP (RSD), and a rapeseed oil denatured with 2% aniline but contained no diesters of PAP (RSA). Female MRL/lpr mice, 7 weeks of age, received orally either an undiluted (neat) or a 1:10 diluted dose of each test oil, canola oil (oil control), water (nai;ve control), or 50-ppm mercury (positive control). Half of each group was sacrificed after 5 weeks of exposure and the remaining mice after 10 weeks of exposure. Serum IgG1, IgG2a, IgE isotypes and antinuclear (ANA), collagen type II, histone, single-stranded DNA (ssDNA), double-stranded DNA (dsDNA) and Sm autoantibody concentrations were determined after 5 and 10 weeks of exposure. The oils did not significantly affect the concentrations of the serum immunoglobulins, although a shift in the IgG1:IgG2a ratio towards IgG1 was noted from 12 to 17 weeks of age (5-10 weeks of treatment). The oils did however stimulate the systemic autoimmune response. The RSD neat treatment resulted in a nonsignificant but noted increase in autoantibodies to collagen (10 weeks), histone (10 weeks) and dsDNA (5 and 10 weeks). CO756 neat increased the serum levels of ANA (5 weeks), collagen (5 weeks) and dsDNA (5 and 10 weeks). The RSA 1:10 dilution increased ssDNA and dsDNA autoantibodies at 5 weeks. The results suggest that PAP is an active principle of these noted responses. These data, coupled with the toxicology and pathology data from this study (Toxicol. Path. 29 (2001) 630), revealed that the three oils incited induction of the lymphoproliferative syndrome and that the two oils containing PAP induced and enhanced the systemic autoimmune response that develops spontaneously at an early age in the MRL/lpr mouse. There was also a positive correlation noted between serum autoantibody concentrations and progression of the idiopathic autoimmune syndrome in the MRL/lpr mouse.