Tryptophan Rich Antigens (TRAgs) are found in all Plasmodium species, but are significantly expanded in number in Plasmodium vivax and related species. The P. vivax genome contains 37 TRAgs, many of which are highly expressed in blood-stage parasites and some were previously localised to the merozoite surface. However their function, and potential role in invasion, is under-explored. We expressed 25 full-length P. vivax TRAgs using the HEK293E cell system, and raised polyclonal antibodies against several to assess their location. Only half of the P. vivax TRAgs are predicted to have N-terminal signal peptides, but our immunofluorescence assays detected surface staining in merozoites from the Plasmodium knowlesi model, irrespective of the predicted signal peptide. We believe there are potential errors in genome annotation and all P. vivax TRAgs are likely secreted. We used purified TRAg proteins in a flow cytometry assay to assess their ability to bind human erythrocytes and reticulocytes. One protein, TRAg25, bound preferentially to reticulocytes. To understand the molecular basis of this binding, we solved the crystal structure of its C-terminal tryptophan rich domain at 1.6Å resolution, which suggested the presence of a potential lipid-binding fold, which is likely conserved across all TRAgs. The ability of recombinant TRAg25 to bind lipids was confirmed using lipid dot blots and liposome binding assays, and provides a potential explanation for the preference of this protein for reticulocytes. A P. knowlesi line with the homologous gene deleted had a reduced invasion efficiency of reticulocytes, supporting a functional role for TRAg25 in parasite invasion. In conclusion, we have reported the first structure of a Plasmodium TRAg, and postulate a potential function for the family in lipid binding and parasite invasion during blood stages. This project is a pilot for a larger systematic reverse vaccinology screen for novel P. vivax blood-stage vaccine candidates.