The autoantibodies found in human and murine systemic lupus erythematosus (SLE) are generally directed against cells or components of cells such as nuclear antigens. This predilection may be due to the unusual immunogenicity of certain autoantigens, or to unusual patterns of antibody crossreactivity. Alternatively, the observed spectrum of reactivities may reflect the in vivo absorption of those autoantibodies directed against soluble antigens. To test whether hitherto undetected autoantibodies against serum proteins might exist in murine SLE, we developed assays that were independent of the possibility of absorption of autoantibodies by serum autoantigens; large numbers of plaque- forming cells (PFC) directed against mouse albumin and mouse transferrin were easily detected in the spleens of MRL/Mp-lpr/lpr, BXSB, and NZB mice. The secreted antibodies were relatively specific for the mouse proteins, since only limited cross-reactivity was seen with albumin and transferrins of other species in inhibition experiments. The production of these hidden antibodies could not be the result of diffuse polyclonal B cell activation, since the PFC to mouse transferrins and albumin were not always accompanied by comparable numbers of PFC against related albumins and transferrins. The results indicate that autoantibody production in murine lupus is a generalized phenomenon, not limited to the production of autoantibodies to nuclear or other cell-bound antibodies. However, the relative specificity of the autoantibodies for self-antigens indicates that diffuse polyclonal B cell activation cannot be the mechanism responsible, and argues that a selective mechanism, probably driven by antigen, accounts for production of autoantibodies in SLE.