Intravenous IgG (IVIG) preparations must meet up with the most stringent standards of safety and quality. Careful donor selection and screening are essential but per se do not guarantee product safety. Before the introduction of anti-HIV and anti-HCV screening, the most important step for the elimination of infected donations, it was likely that infectious virus was present in the plasma pools used to make IVIG. However, most of these plasma products have never transmitted any virus diseases and show a good record of viral safety. This strongly suggests that certain steps utilised in the manufacturing of IVIG have the capacity of reducing the initial virus load by elimination, by inactivation or by both mechanisms simultaneously. In this study, we have validated several steps in the manufacturing process of Sandoglobulin in order to establish the mechanisms responsible for the viral safety of this product. The individual steps in the manufacturing process were scaled down to laboratory level and each step was challenged with the model viruses HIV. Semliki Forest virus (SFV), Sindbis virus (SV), pseudorablies virus (PRV), bovine viral diarrhea virus (BVDV) or bovine enterovirus (BEV) independently. The filtration steps in the presence of filter aids with and without cold ethanol precipitation are extremely efficient in removing the viruses. Typically, a single step exhibit about 3 to 5 logs virus reduction resulting in an overall removal of 12 to 16 logs for the fractionation process. In addition to the fractionation, the manufacturing process of this IVIG product includes treatment of the IgG solution at pH 4, 37 °C in the presence of trace amounts of pepsin. This step was analysed in order to determine the inactivation kinetics. This treatment exhibits and efficient and rapid inactivation of HIV, SFV, BVDV and PRV yielding an additional reduction of 4 to 7 logs of the respective viruses. We have further examined the influence of various parameters on the rate of virus inactivation during the pH 4/pepsin treatment by changing the ionic strength, sucrose and protein concentrations of the IgG solution. We found that changing these conditions influences the rate of virus inactivation. In conclusion, the reduction factors for viral inactivation by pH 4/pepsin treatment and by fractionation and filtration demonstrate that the process leads to a total reduction in viral load as high as 21 logs and explains the excellent safety record of this specific IVIG preparation. Recently, we started to investigate the efficacy on virus removal by nanofiltration of IgG solutions. This technology might be introduced in the future to further increase the safety margin of IgG preparations. Results obtained on a laboratory scale will be discussed.