The importance of feedstuff evaluation as part of ration formulation is recognised by UK farmers, with the number of silage samples analysed increasing annually. These can be undertaken by the UK advisory services or feed companies, with several commercial laboratories offering similar services. However, as discussed by Beever [1993], there is concern over the authenticity of some estimates of feeding value, in relation to the analytical techniques used, the lack of agreed standardised procedures, and the apparent variation in results which exists between laboratories. This study examined the extent of this variation when 2 maize silage samples were independently analysed by 9 different laboratories, and compared laboratory based estimates of metabolisable energy [ME] contents with those derived by feeding the same diets to lactating cows.