Milk antibiotic residues have been a public concern in recent years. The Grade A Pasteurized Milk Ordinance mandates that raw Grade A milk will test negative for beta-lactam antibiotic residues before processing. The purpose of this research was to investigate the ability of various levels of peroxide and heat to inactivate penicillin G in raw milk. Whole milk spiked to a mean of 436 +/- 15.1 (standard error of the mean) ppb of potassium penicillin G was treated with hydrogen peroxide at levels of 0.0, 0.09, 0.17, and 0.34%. Samples at each peroxide level (n = 6 per treatment) were treated as follows: 1) incubated at 54.4 degrees C for 3 h, 2) pasteurized at 62.8 degrees C for 30 min, 3) incubated and pasteurized as in treatments 1 and 2, or 4) received no further treatment. A beta-lactam competitive microbial receptor assay was used for quantification of penicillin G. Concentrations of penicillin in selected samples were determined by HPLC for a comparison of test methods. Treatments were evaluated relative to their ability to reduce milk penicillin G levels to below the safe level of 5 ppb. The 0.09% hydrogen peroxide level was ineffective for all treatments. Hydrogen peroxide at 0.17% lowered the mean penicillin G (+/- SEM) from 436 +/- 15.1 to 6 +/- 1.49 ppb using the incubated and pasteurized heat treatment. The 0.34% concentration of hydrogen peroxide was the most effective, inactivating penicillin G to a level well below the safe level of 5 ppb with the pasteurized heat treatment, with or without incubation.