A current hypothesis explaining the aging process implicates the accumulation of oxidized protein in animal tissues. This hypothesis is based on a series of reports showing an age-dependent increase in protein carbonyl content and an age-dependent loss of enzyme function. This hypothesis is also supported by the report of a novel effect of N-tert-butyl-alpha-phenylnitrone (PBN) in reversing these age-dependent changes. Here we specifically study the method that was used to measure reactive protein carbonyls in tissues. This method uses 2,4-dinitrophenylhydrazine (DNPH) and includes a washing procedure. Our results indicate that reactive protein carbonyls in normal crude tissue extracts cannot be reliably measured by this method, although it does reliably measure reactive carbonyls in purified proteins which have been oxidatively modified in vitro. The nucleic acids in tissues could be a major problem encountered in the assay. Using the streptomycin sulfate treatment combined with a dialysis step, we were successful in removing most nucleic acids from a crude tissue extract, but then the reactive carbonyl level in the crude tissue extract was too low to be reliably measured. This streptomycin sulfate treatment procedure, however, had no effect on the reactive carbonyl measurement of an oxidized protein sample. The unwashed free DNPH was another major problem in the assay because of its very strong absorption around 370 nm, where reactive carbonyls were quantitated. Nevertheless, on using the procedure described in the literature to measure total "reactive carbonyls" in rat liver and gerbil brain cortex, no change with age or PBN treatment was found. Then, we investigated a HPLC procedure which uses sodium dodecyl sulfate in the mobile phase but this was also found to be unsuitable for the reactive protein carbonyl assay in tissues.