The effect of blind passage and centrifugation on the isolation of bovine coronavirus in human rectal tumor cells cultured in shell vials was investigated. A total of 68 fecal samples known to be positive for bovine coronavirus by transmission electron microscopic (TEM) examination were used. The samples were centrifuged onto human rectal tumor cell monolayers and incubated in the presence of trypsin. The growth of bovine coronavirus in infected cells was demonstrated by fluorescent antibody staining, and the extracellular virus was detected and confirmed by hemagglutination and hemagglutination-inhibition tests, respectively. Of the 68 TEM-positive samples, 51 (75%), 58 (85%), and 61 (90%) grew in shell vial cell cultures at first, second, and third passages, respectively. Of the 51 cultures positive on first passage, 19 were examined by TEM; 18 of these were positive for bovine coronavirus. The shell vial technique was also compared with direct detection of bovine coronavirus by staining cryostat sections of infected tissues in a direct fluorescent antibody assay. The results of direct fluorescent antibody assay were available for 54 of the 68 samples, of which 53 (98%) and 43 (80%) were positive by shell vial technique and direct fluorescent antibody assay, respectively. For identification of bovine coronavirus, shell vials using human rectal tumor cells in the presence of trypsin is more sensitive than direct fluorescent antibody assay but is relatively less sensitive than transmission electron microscopy.