The safety and efficacy of biopharmaceuticals can be severely impaired by their immunogenicity. A risk-based strategy should be used to assess immunogenicity on a case-by-case basis using standardized methods to correlate anti-drug antibody levels with clinical outcome. In silico and in vitro techniques allow putative T-cell epitopes to be identified and eliminated in candidate molecules while maintaining structure and function. Putative T-cell epitopes can be studied in the context of the HLA allotypes representative of the target population in vitro and in transgenic mice that express human HLA genes. Mice immune tolerant to human proteins allow the study of the effect of factors such as aggregation on the loss of immune tolerance. However, significant challenges remain in order to be able predict the immunogenicity of a therapeutic protein in a particular individual.