Five microbiological methods for detecting residues of penicillin and penicillin-like antibiotics were compared, and the relative incidence of residues in whole milk samples was determined. Five groups of 10 lactating cows each and five commercial antibiotic products (penicillin G in oil, penicillin G in a water-dispersible oil base, cloxacillin, cephapirin and ampicillin) used for the treatment of mastitis were studied. Milk samples collected from the cows were tested for antibiotic content by the Sarcina lutea or Bacillus subtilis cylinder plate, International Dairy Federation-Association of Official Analytical Chemists (IDF-AOAC) Bacillus stearothermophilus ATCC 10149 disc plate, Delvotest-P, Antibiotic Residue Branch (ARB) B. stearothermophilus C 953 quantitative disc plate and ARB B. stearothermophilus C 953 quantitative cylinder plate methods. All samples tested by the original S. lutea or B. subtilis cylinder plate methods, with the exception of the penicillin G in oil product (Uni-Biotic), were free of antibiotic residues and showed no inhibitors beyond the labeled milk withdrawal time. The newer, more sensitive official IDF-AOAC method, Delvotest-P and both ARB B. stearothermophilus quantitative disc and cylinder plate methods detected measurable amounts of cephapirin and penicillin G in the post-treatment milk for one to three additional milkings beyond the labeled milk discard time. Cloxacillin, as determined by the ARB B. stearothermophilus cylinder plate method, was generally depleted by 48 h. The excretion rate of ampicillin product coincided with the labeled milk discard time. The level of detection of both disc plate methods and Delvotest-P for cephapirin and ampicillin was comparable. Variations in the concentration in milk and the persistence of two penicillin G products were probably due to differences in the formulation and the dosing regimen. Inhibitory substances and non-specific reactions were observed with the IDF-AOAC disc method, the Delvotest-P and both ARB B. stearothermophilus methods in some control and post-treatment milk samples. These results are consistent with reports of other investigators and suggest that the use of more sensitive procedures for monitoring antibiotic residues in milk may result, for some infusion preparations, in an increased incidence of violative residues of penicillin G and cephapirin.