30 years ago, the potential of bispecific antibodies to engage cytotoxic T cells for the lysis of cancer cells has been discovered. Today a variety of bispecific antibodies against diverse cell surface structures have been developed, the majority of them are produced in mammalian cell culture systems. Moreover, transgenic animals can be used to express bispecific antibodies. This is the case for r28M. This tandem single-chain variable fragment antibody is designed to target human CD28 and the melanoma/glioblastoma-associated cell surface chondroitin sulfate proteoglycan 4. Besides the r28M, described here, no such bispecific antibody is known to be expressed by transgenic livestock. In this study we investigated the large-scale purification and biological activity of the bispecific antibody r28M, expressed in the blood of transgenic cattle. We revealed the most effective bovine blood collection procedure and developed a competitive ELISA to quantify r28M in the blood of the transgenic animals as well as in purified form. Subsequently, we optimized the purification protocol for large-scale purposes and separated the resulting r28M-enriched fraction by size exclusion chromatography into monomers, dimers and aggregates. Cell culture experiments regarding the biological activity revealed the monomer fraction as being the most appropriate one to continue working with. Besides the detailed characterization of the antibody’s activity, first insights into tumor cell death pathways mediated by r28M-activated peripheral blood mononuclear cells were gained. In consideration of possible applications in vivo we also tested the effect of the addition of different excipients to r28M. Summing up, we managed to develop a “stable to lab”-large-scale purification system of monomeric r28M out of bovine plasma. Further we could prove that the antibody’s biological activity is unaffected and still highly specific and thus, might be applicable for the treatment of melanoma.