An ELISA has been developed for measurement of milk and serum IgG concentrations directed against Salmonella dublin. Four groups of cows were studied: group A--7 experimentally challenge-exposed cows (infected, recovered group); group B--6 normal uninfected randomly selected control cows; group C--7 naturally occurring S dublin carrier cows; and group D--6 normal uninfected S dublin negative cows from the same herd as group C. Group-A cows were inoculated orally, or inoculated orally and then IV, but none became a S dublin carrier. As expected, all 7 group-A cows responded with a marked increase in ELISA titer after oral exposure to virulent S dublin, starting with a mean serum titer of 17.7% and reaching a peak mean serum titer of 79.3% approximately 76 days after initial exposure. As determined by necropsy and organ culturing of the remaining cows, none of the group-A cows became carriers. The mean serum ELISA titer for group-B uninfected control cows was 14.1% (SD +/- 12.8%). The mean milk ELISA titer was -1.0% (SD +/- 5.5%). Colostrum and then milk gave false-positive results for up to 2 weeks after onset of lactation. Group-B cows were culture negative for S dublin in feces and milk during lactation, and when tissues were cultured after euthanasia. Milk and serum samples for ELISA, and milk and fecal samples for culturing were taken from all group-A and -B cows twice a week for 6 months. Statistical correlation (P less than 0.05) was found between serum and milk ELISA titers.(ABSTRACT TRUNCATED AT 250 WORDS)