Birds are most susceptible to infection by avian reovirus, genus Orthoreovirus family Reoviridae, at a young age. Although chicks are protected by antibodies transferred from vaccinated maternal flocks, due to the many variants in the field, the efficiency of the vaccines is limited. The level of antibodies against viruses is generally determined by enzyme-linked immunosorbent assay (ELISA), using the whole virus as the antigen. This has some disadvantages: first, the test measures antibodies against all capsid proteins, most of which are irrelevant for neutralizing the virus, and as such does not reflect the real protection status; second, it is impossible to distinguish between vaccine- and infection-derived antibodies. In the case of a virus that changes frequently, a third disadvantage is the inability to distinguish among serotypes. The aim of this study was to develop a test that would address these concerns. Four prototypes of the avian reovirus protein sigma C were used as antigens on the ELISA plate. Sigma C is the main protein inducing neutralizing antibodies and the most variable among strains and isolates, and it is used for reovirus classification. This differentiating ELISA enabled distinguishing between vaccine and field strains of the virus, identifying the infection source, and in the case of vaccination, exclusively determining the level of protective antibodies. Whereas the whole virus detected antibodies against all strains, differentiating ELISA enabled differentiating between infected and vaccinated animals (DIVA) and in most cases, identifying the sigma C genotype. In a field study, a correlation was found between disease symptoms and antibodies identified against virulent strains in the flock. Thus virulent strains can be identified in the field, enabling adjustment of the relevant vaccines.