The performances of 2 official methods for iodine analysis based on inductively coupled plasma-mass spectrometry (ICP-MS) and the ion-selective electrode (ISE) method were compared for milk-based products. The aim of the study was to determine the performance characteristics of both methods to check the labeled concentration of iodine. Good precision was found for both methods with highest relative standard deviation of repeatability (RSD(r)) at 2.3 and 2.7% for ISE and ICP-MS, respectively. Intermediate reproducibility (RSD(iR)), single laboratory within 6 different days, was also good with the highest values at 7.3 and 8% by ISE and ICP-MS, respectively. Measurement uncertainty was estimated based on the RSD(iR) data, and it was concluded that both methods were capable of determining iodine concentrations within an uncertainty below +/- 20%. The accuracy of the methods was determined by analyzing certified reference materials, in-house proficiency test samples, and commercial products. Both methods returned similar results when applied on freshly opened samples. In samples that had been opened and kept exposed to air during storage, ISE returned lower iodine concentrations than ICP-MS. In commercial samples, the linear regression between both methods was ISE = 0.95 x ICP-MS -0.060 for freshly opened samples and ISE = 0.85 x ICP-MS + 0.069 for samples exposed to air. The tendency of ISE to return lower results than ICP-MS is explained by the fact that ISE is sensitive to iodide but does not measure iodine that may be bound organically to the matrix. This seems to be more pronounced in samples that were stored longer. Because in most countries iodine is labeled as total iodine, acceptance of an international standard based on the ICP-MS technique which takes all forms of iodine into account, is recommended. This would help to avoid any potential dispute on the accuracy of labeled iodine concentrations in finished products.