As a consequence of increased insight into the cellular and molecular mechanisms responsible for induction of B cell and T cell autoimmunity to DNA and nucleosomes, there is an obvious need to reconsider the dogma stating that anti-dsDNA antibodies serve as marker antibodies for SLE and also that anti-dsDNA antibodies per se are responsible for the initiation of lupus nephritis. Given that the potential to produce anti-dsDNA antibodies is an inherent property of the normal immune system and that few anti-DNA antibodies have nephritogenic potential, we must try to solve the problem whether it is avidity for DNA, specificity for unique DNA structures or cross-reactivity with non-DNA molecules, that make such antibodies pathogenic and thus potential markers for SLE and lupus nephritis. In this review, we will summarize contemporary problems related to these questions; (1) try to focus on phenotypic differences with respect to the ability to produce anti-dsDNA antibodies between individuals suffering from SLE and those not belonging to this diagnostic group, and (2) to describe differences between pathogenic and non-pathogenic anti-dsDNA antibodies.