The mechanism of action of intravenous immune globulin (IVIG) in the therapy of autoimmune disease has been speculated upon for many years. Previous studies have raised the possibility that IVIG acts via an effect on IgG Fc receptors (FcRs) on phagocytic cells and B lymphocytes, on the production of anti-idiotype antibody, or on the control of the immune response. In the course of our studies of complement function we were struck by the fact that complement activation often leads to the binding of complement components to individual immunoglobulin molecules. For example, C3 has been shown to bind to the Fd fragment of IgG in the form of a C3b/IgG one-to-one complex. The C3b/IgG complex has new properties that differ from those of either IgG or C3b alone in that the complex can interact with two receptors on phagocytes: CR1, which recognizes C3b, and FcR, which recognizes the Fc fragment of IgG. Particles opsonized with IgG/C3b interact with both receptors and are phagocytized rapidly. The complex acts as a superopsonin and superlysin. IgG/C3b resists degradation by factors H and I, which also adds to its inflammatory potential. We and others have noted that bacteria coated with immunoglobulin and then incubated in serum have C3 deposited on their surfaces, which in many instances is bound to the IgG molecules. For example, we found that 30% of the C3 deposited on antibody-coated pneumococci is bound not to the pneumococcal surface but rather to the coating immunoglobulin. We reasoned that IVIG may act as a receptor for activated complement components, preventing their attachment to targets. This was tested directly in a number of animal and human models. The results of these tests and their clinical implications are presented.