It has been proposed that the "normal" stimulation of the immune system that occurs from interactions with environmental stimuli, whether infectious or dietary, is necessary for the initiation and/or continuation of autoimmunity. We tested this hypothesis by deriving a group of MRL-lpr mice into a germfree (GF) environment. At 5 mo of age, no differences between GF and conventional MRL-lpr mice were noted in lymphoproliferation, flow cytometric analysis of lymph node cells (LN), or histologic analysis of the kidneys. Autoantibody levels were comparably elevated in both groups. A second experiment tested the role of residual environmental stimuli by contrasting GF mice fed either a low m.w., ultrafiltered Ag-free (GF-AF) diet or an autoclaved natural ingredient diet (GF-NI). At 4 mo of age, both groups showed extensive lymphoproliferation and aberrant T cell formation, although the GF-AF mice had approximately 50% smaller LNs compared with sex-matched GF-NI controls. Autoantibody formation was present in both groups. Histologic analysis of the kidneys revealed that GF-AF mice had much lower levels of nephritis, while immunofluorescence analysis demonstrated no difference in Ig deposits but did reveal a paucity of C3 deposition in the kidneys of GF-AF mice. These data do not support a role for infectious agents in the induction of lymphoproliferation and B cell autoimmunity in MRL-lpr mice. Furthermore, they suggest that autoantibodies do not originate from B cells that were initially committed to exogenous Ags. They do suggest a possible contributory role for dietary exposure in the extent of lymphoproliferation and development of nephritis in this strain.