No standardized method exists for the determination of total or individual carbohydrates in beer and this could cause difficulties in view of regulatory requirements for labelling. A comparative study of known methods likely to be most suitable for the measurement of total carbohydrate in beer was therefore undertaken. Methods were also examined for estimating other important residual carbohydrates in beer, such as residual fermentable carbohydrates, β-glucan, total fructose and fructosans, and total pentose and pentosans. The beers analysed covered a wide range of carbohydrate levels and, as expected, the different methods gave different values for the carbohydrate content of the same beer. We consider that the colorimetric anthrone-sulphuric acid procedure (Yadav's version) is to be preferred for estimating the total residual carbohydrate content of beer. It is less prone to interference, relatively simple, rapid and gives more accurate results than acid or enzymic hydrolysis in combination with reductometric procedures. The within-laboratory coefficient of variation is 1.68%.