A collaborative study was conducted of the Food and Drug Administration (FDA)-optimized Monier-Williams method for determining sulfites in foods. Twenty-one industry and government laboratories participated in the study, which was jointly sponsored by the National Food Processors Association and FDA. Familiarization samples were shipped to each collaborator. Collaborators were permitted to proceed to the main study only after they demonstrated ability to perform the method to ensure that the study tested the performance of the method itself and not that of the individual laboratories. The study design involved 3 food matrixes (hominy, fruit juice, and protein [seafood]). Each matrix was prepared at 3 sulfite levels--the regulatory level, half the regulatory level, twice the regulatory level--and as a blank. All test samples were analyzed as blind duplicates, which gave each collaborator a total of 24 test portions. Collaborative recoveries gave a reproducibility (among-laboratories) coefficient of variation that ranged from 15.5 to 26.6% for sulfite determined as SO2 by weight in the 3 foods at the 10 ppm level. The optimized Monier-Williams method has been approved interim official first action to replace the AOAC modified Monier-Williams method, 20.123-20.125.