Changes in the milk antibody levels against Staphylococcus aureus were measured at the start of an experimental intramammary instillation of either S. aureus (Study I) or Staphylococcus hyicus (Study II). A commercial enzyme-linked immunosorbent assay system was used. Twenty-one Holstein cows were enrolled in Study I and 15 Holstein cows were used in Study II. Pathogen instillation began 21 days before the start of the non-lactating period. Cows received intramammary antibiotic treatment in all quarters immediately after the last milking, the start of the non-lactating period. Lacteal secretions were collected before the start of the non-lactating period, and during the immediate postpartum period in both studies, and during the non-lactating period in Study I. Milk was cultured for mastitis pathogens and S. aureus antibody levels and somatic cell counts were determined from all samples. There was an approximate 2-week delay in the elevation in antibody levels in response to the instillation of S. aureus. Antibody levels remained elevated in cows with S. aureus intramammary infections postpartum, but were below threshold in cows where intramammary infections were cured during the non-lactating period. Antibody levels were elevated by S. hyicus intramammary infections, remained elevated for the first 12 days postpartum, but were below threshold by day 21 postpartum. Cows with incipient intramammary S. aureus infections might be misclassified as false negatives by the antibody test. However, results suggest that cows with S. hyicus intramammary infections that were not cured would not be misclassified if milk is withheld from test for the first 30 days postpartum, as recommended by the manufacturer of the test.