Initial food industry testing in our laboratory using enzyme-linked immunosorbent assay (ELISA) methods indicated that the darkest caramel color (class IV) unexpectedly contained traces of peanut protein, a potential undeclared allergen issue. Caramel production centers on the heating of sugars, often glucose, under controlled heat and chemical processing conditions with other ingredients including ammonia, sulfite, and/or alkali salts. These ingredients should not contain any traces of peanut residue. We sought to determine the reliability of commercially available peanut allergen ELISA methods for detection of apparent peanut residue in caramel coloring. Caramel color samples of classes I, II, III, and IV were obtained from 2 commercial suppliers and tested using 6 commercially available quantitative and qualitative peanut ELISA kits. Five lots of class IV caramel color were spiked with a known concentration of peanut protein from light roasted peanut flour to assess recovery of peanut residue using a spike and recovery protocol with either 15 ppm or 100 ppm peanut protein on a kit-specific basis. A false positive detection of peanut protein was found in class IV caramel colors with a range of 1.2 to 17.6 parts per million recovered in both spiked and unspiked liquid caramel color samples. ELISA kit spike/recovery results indicate that false negative results might also be obtained if peanut contamination were ever to actually exist in class IV caramel color. Manufacturers of peanut-free products often test all ingredients for peanut allergen residues using commercial ELISA kits. ELISA methods are not reliable for the detection of peanut in class IV caramel ingredients and their use is not recommended with this matrix. [ABSTRACT FROM AUTHOR]