The objective of this study was to evaluate the effects of colostrum and transition milk composition on specificity rates of antibiotic residue screening tests. Milk from 25 primigravid Holstein heifers was collected from either first, second, or third milking (colostrum) and from either fifth, sixth, or seventh milking (transition milk) following parturition. Milk sampled was visibly normal and heifers were not treated with an antibiotic within 30 d before parturition. Quarter foremilk samples were collected aseptically and analyzed for mastitis pathogens. A sample from the total composite milk was analyzed for somatic cell counts (SCC), milk protein and fat, immunoglobulin concentrations and for antibiotics using four antibiotic residue screening tests. Mastitis pathogens were present in colostrum from 36% of heifers (n = 9) and from 16% of heifers (n = 4) in the subsequent transition milk. Mean SCC were 2,458,000 and 866,000 counts/ml and IgG1 concentrations were 22.7 and 3.07 mg/ml for colostrum and transition milk, respectively. Specificity rates of the screening tests ranged from 0.16 to 0.88 for colostrum and 0.60 to 1.0 for transition milk. Increased milk protein and IgG1 concentrations in milk were associated with an increase in the probability of a false positive outcome for the Charm Cowside (Charm Sciences, Inc., Malden, MA), CITE Snap (IDEXX Laboratories, Westbrook, ME), and Penzyme (Cultor Food Science, Milwaukee, WI) tests. Fat content of milk was positively related to an increase in false positive rates for the CITE Snap test. Milk should not be tested for antibiotic residues before the sixth milking after parturition to avoid high rates of false positive outcomes.