Despite a good association between anti-DNA autoantibodies and lupus nephritis, it is difficult to determine the pathogenic potential of an anti-DNA autoantibody response. It is proposed that anti-DNA antibodies can exert their pathogenic effects through deposition as immune complexes in the kidney or through the recognition of cross-reactive antigens in the kidneys. Several studies in literature demonstrate that cross-reactivity of anti-DNA antibodies with kidney antigens is critical for their pathogenic potential. This raises the question whether DNA is responsible for the activation and selection of B cells generating cross-reactive anti-DNA antibodies. Recent studies suggest that antigens other than DNA can initiate an antibody response that is cross-reactive with dsDNA. Moreover, we and other have demonstrated that lupus nephritis can occur in the absence of anti-DNA antibodies. Thus, reactivity to dsDNA should be considered as one of the characteristic of polyreactive autoantibodies and not a primary requisite for the pathogenesis of lupus nephritis.