An optimized DNA extraction protocol for animal tissues coupled with sensitive PCR methods was used to determine whether trace levels of feed-derived DNA fragments, plant and/or transgenic, are detectable in animal tissue samples including dairy milk and samples of muscle (meat) from chickens, swine, and beef steers. Assays were developed to detect DNA fragments of both the high copy number chloroplast-encoded maize rubisco gene (rbcL) and single copy nuclear-encoded transgenic elements (p35S and a MON 810-specific gene fragment). The specificities of the two rbcL PCR assays and two transgenic DNA PCR assays were established by testing against a range of conventional plant species and genetically modified maize crops. The sensitivities of the two rbcL PCR assays (resulting in 173 and 500 bp amplicons) were similar, detecting as little as 0.08 and 0.02 genomic equivalents, respectively. The sensitivities of the p35S and MON 810 PCR assays were approximately 5 and 10 genomic equivalents for 123 bp and 149 bp amplicons, respectively, which were considerably less than the sensitivity of the rbcL assays in terms of plant cell equivalents, but approximately similar when the higher numbers of copies of the chloroplast genome per cell are taken into account. The 173 bp rbcL assay detected the target plant chloroplast DNA fragment in 5%, 15%, and 53% of the muscle samples from beef steers, broiler chickens, and swine, respectively, and in 86% of the milk samples from dairy cows. Reanalysis of new aliquots of 31 of the pork samples that were positive in the 173 bp rbcL PCR showed that 58% of these samples were reproducibly positive in this same PCR assay. The 500 bp rbcL assay detected DNA fragments in 43% of the swine muscle samples and 79% of the milk samples. By comparison, no statistically significant detections of transgenic DNA fragments by the p35S PCR assay occurred with any of these animal tissue samples.