Urine specimens submitted for the diagnosis of cytomegalovirus infection were inoculated into shell vials that had been pretreated with a combination of dimethyl sulfoxide (DMSO) and dexamethasone (DEX). The results were compared with those for inoculated shell vials which had received no drug treatment. Of 664 specimens, 100 (15%) were positive for cytomegalovirus. Of the 100 strains of cytomegalovirus, 88 (88%) were detected in both DMSO-DEX-treated and untreated shell vials. Of the remaining 12 positive specimens, 6 were detected with untreated shell vials exclusively and 6 were detected with DMSO-DEX-treated shell vials alone (not significant by the sign test). The median number of fluorescent foci was not significantly higher in DMSO-DEX-treated shell vials compared with that in untreated cultures (Wilcoxon signed-rank test; P = 0.1). DMSO-DEX-treated monolayers did not enhance the sensitivity detection of cytomegalovirus in shell vial cell cultures.