Immunogenicity assessment is an important issue for ensuring the safety and efficacy of therapeutic protein products. Although the reliability of the anti-drug antibody (ADA) assay is one of the key points, there are some difficulties in assessing its validity because the analytes are polyclonal antibodies with variable and unknown characteristics. To elucidate the points to consider for the ADA assay, a Japanese research group was established that discusses the issues raised on the immunogenicity assessment. In this review, we first introduce the current situation regarding the development and immunogenicity assessment of therapeutic protein products in Japan. We then present our current view and recommendations on the ADA assay by considering its unique features. Therapeutic proteins such as monoclonal antibodies are currently essential in the treatment of cancer, autoimmune disease and other diseases. Many therapeutic protein products intended to address unmet medical needs are being developed worldwide, including in Japan, and therefore, ensuring their safety and efficacy is of paramount importance. Since protein has its intrinsic feature of immunogenicity owing to its structure containing potential B-cell and T-cell epitopes, therapeutic proteins have the potential to induce ADA even if the protein has the same amino acid sequence as endogenous human proteins. Proteins can be recognized as antigens by B cells and are also incorporated and digested by antigen-presenting cells. This process leads to the production of T-cell epitope peptides that are presented on major histocompatibility (MHC) II molecules. Various patient-and product-specific factors are suggested to affect the process of ADA induction [1]. The emergence of ADA in patients can potentially lead to loss of efficacy and/or adverse events. Therefore, immunogenicity risk assessment and risk-mitigating strategies are required during the development of therapeutic protein products [2].The appropriateness of the ADA assay is a key issue in immunogenicity assessment. The ligand-binding assay (LBA) is often used for detecting ADA in biological samples. Unlike the assay used for drug concentration analysis, the ADA assay has unique features, for example: the actual characteristics of the analyte (human ADA) are variable and unknown; there is no real reference standard: the experimentally prepared positive control for ADA is used as a surrogate reference standard to evaluate and control the assay performance; and the evaluated validation parameters such as sensitivity and drug tolerance limit (DTL) vary depending on the characteristics, mainly the affinity of the used positive control. In addition, matrix components that interfere with the assay in the study samples may vary depending on the disease and/or the individual patients. Therefore, it is practically impossible to completely assess the validity of the ADA assay, and the assay is associated with certain risks of obtaining an inappropriate result. As mentioned in the European Medicines Agency (EMA...