Scientific relevance. Medicinal products based on immunoglobulin class G (IgG) from human plasma are widely used in clinical practice to treat bacterial and viral infections, primary and secondary immunodeficiencies, and autoimmune diseases. Nanofiltration is a way to mitigate the risk of in-process contamination of raw materials with various pathogens, including viruses. Therefore, it is relevant to investigate the development and implementation of additional viral inactivation and/or elimination steps.Aim. This study aimed to develop and validate optimum nanofiltration conditions and to scale up the nanofiltration step for the manufacturing of human IgG for intravenous administration.Materials and methods. The study used a solution of IgG from plasma fractions II and III. The authors paired nanofilters manufactured by Planova 20N and BioEx (Asahi Kasei, Japan), Viresolve Pro (Merck Millipore, USA), Virosart HC and HF (Sartorius, Germany), and Pegasus SV4 and Prime (Pall, USA) with Sartopore polyethersulphone prefilters by Sartorius (Germany), Virosart MAX polyamide prefilters by Sartorius (Germany), and EKX-P regenerated cellulose prefilters by Pall (Germany). Virus reduction validation studies were performed with model viruses (human immunodeficiency virus type 1, porcine transmissible gastroenteritis virus, porcine parvovirus, murine encephalomyocarditis virus, and bovine viral diarrhoea virus) in the laboratories of the N.F. Gamaleya centre. The sample data analysis involved calculating mean values with 95% confidence intervals.Results. For all the selected combinations of prefilters and filters, the maximum nanofiltration throughput depended on the IgG concentration in the test solution. With the combination of an EKX-P filter with a Pegasus SV4 nanofilter, the maximum throughput and the IgG yield reached 6300 g/m2 and 95%, respectively. When combined with a Planova 20N nanofilter, EKX-P and Sartopore (polyethersulphone) filters provided a maximum throughput of up to 2980 g/m2 and an IgG yield of almost 100%, provided that the test solution had an IgG concentration of 10 g/L. With different filter combinations, virus reduction levels ranged from 4.00±0.05 to 4.75±0.04 log10 for human immunodeficiency virus type 1, from 4.30±0.04 to 4.55±0.06 log10 for porcine transmissible gastroenteritis virus, from 5.38±0.08 log10 to 5.57±0.04 log10 for murine encephalomyocarditis virus, 5.12±0.10 log10 to 5.25±0.08 log10 for porcine parvovirus, and exceeded 5.00 log10 for bovine viral diarrhoea virus. The virus reduction levels achieved were not statistically associated with prefilter brands.Conclusions. The study demonstrated that nanofiltration was effective at removing viruses with various virion sizes and physicochemical characteristics, including viruses as small as parvovirus B19. The levels of virus reduction exceeded 4 log10 and met the acceptance criteria.The laboratory-scale nanofiltration parameters and the corresponding filtration times, as well as IgG yields, did not change when the process was scaled up. Therefore, nanofiltration is an effective and productive technique that helps eliminate various types of viruses and considerably improve viral safety without affecting the quality of biological medicinal products.