Abstract Analysis for vitamins is difficult and many factors affect data quality. Factors include the matrix, whether the vitamin is naturally present within the matrix or is added, the molecular form of the vitamin, vitamin stability and solubility. Uniformity of the vitamin within the sample is also known to affect results in addition to the methodology used. Proficiency testing is an established quality assurance measure and is based on comparison of results of laboratories in an interlaboratory trial [Thompson, M., Ellison, S.L.R., & Wood, R. (2006). The international harmonised protocol for the proficiency testing of analytical chemistry laboratories. Pure and Applied Chemistry, 78 (1), 145–196]. It highlights problems in laboratory analysis and can be used as an educational tool to help to improve data quality. FAPAS ® proficiency tests 2139 [FAPAS ® (2006a). FAPAS ® proficiency test 2139: Vitamins in liquid supplement: April–June 2006 . York, UK: Central Science Laboratory] and 2141 [FAPAS ® (2006b). FAPAS ® proficiency test 2141: Vitamins in breakfast cereal: August–September 2006 . York, UK: Central Science Laboratory] demonstrate this for the analysis of vitamin B 2 . From the distribution of participants’ results, the lower of the two modes or sub-populations identified for the analysis of liquid vitamin supplement is thought to be due to the incomplete conversion of riboflavin-5-phosphate to riboflavin.