Protein carbonyl content is widely used as both a marker for oxidative stress and a measure of oxidative damage. Widely used methods for determination of protein carbonylation utilize the reaction of carbonyl groups with 2,4-dinitrophenylhydrazine (DNPH) to form protein-bound 2,4-dinitrophenylhydrazones. Hydrazones can be quantitated spectrophotometrically or, for greater sensitivity, detected immunochemically with anti-dinitrophenyl antibodies. Attention to methodology is important to avoid artifactual elevation in protein carbonyl measurements. We studied extracts of Escherichia coli to identify and eliminate such effects. Nucleic acid contamination caused serious artifactual increases in the protein carbonyl content determined by spectrophotometric techniques. Both in vitro synthesized DNA oligonucleotides and purified chromosomal DNA reacted strongly with 2,4-DNPH. Treatment of cell extracts with DNase+RNase or with streptomycin sulfate to precipitate nucleic acids dramatically reduced the apparent carbonyl, while exposure to proteinase K did not. The commercial kit for immunochemical detection of protein carbonylation (OxyBlot from Chemicon/Millipore) recommends a high concentration of thiol in the homogenizing buffer. We found this recommendation leads to an artifactual doubling of the protein carbonyl, perhaps due to a thiol-stimulated Fenton reaction. Avoiding oxidizing conditions, removal of nucleic acids, and prompt assay of samples can prevent artifactual effects on protein carbonyl measurements.