The spike (S) glycoprotein of the Miller strain of transmissible gastroenteritis virus (TGEV) was recently cloned and expressed in baculovirus. The recombinant S protein was used as the coating antigen in a competition (blocking) enzyme-linked immunosorbent assay (ELISA) in combination with monoclonal antibodies to the S protein epitope A (conserved on TGEV and porcine respiratory coronavirus [PRCV]) or epitope D (present on TGEV only) to differentiate PRCV- from TGEV-induced antibodies. One set (set A) of 125 serum samples were collected at different times after inoculation of caesarean-derived, colostrum-deprived (n = 52) and conventional young pigs (n = 73) with 1 of the 2 porcine coronaviruses or uninoculated negative controls (TGEV/PRCV/negative = 75/30/20). A second set (set B) of 63 serum samples originated from adult sows inoculated with PRCV and the recombinant TGEV S protein or with mock-protein control and then exposed to virulent TGEV after challenge of their litters. Sera from set A were used to assess the accuracy indicators (sensitivity, specificity, accuracy) of the fixed-cell blocking ELISA, which uses swine testicular cells infected with the M6 strain of TGEV as the antigen source (ELISA 1) and the newly developed ELISA based on the recombinant S protein as antigen (ELISA 2). The sera from set B (adults) were tested for comparison. The plaque reduction virus neutralization test was used as a confirmatory test for the presence of antibodies to TGEV/PRCV in the test sera. The accuracy indicators for both ELISAs suggest that differential diagnosis can be of practical use at least 3 weeks after inoculation by testing the dual (acute/convalescent) samples from each individual in conjunction with another confirmatory (virus neutralization) antibody assay to provide valid and complete differentiation information. Moreover, whereas ELISA 1 had 10-20% false positive results to epitope D for PRCV-infected pigs (set A samples), no false-positive results to epitope D occurred using ELISA 2, indicating its greater specificity. The progression of seroresponses to the TGEV S protein epitopes A or D, as measured by the 2 ELISAs, was similar for both sets (A and B) of samples. Differentiation between TGEV and PRCV antibodies (based on seroresponses to epitope D) was consistently measured after the third week of inoculation.