Summary A method is proposed for estimating the percentage of dietary protein that is degraded by microbial action in the rumen when protein supplement is added to a specified ration. The potential degradability, p , is measured by incubating the supplement in artificial-fibre bags in the rumen and is related to incubation time, t , by the equation p = a+b (1 – e -ct ). The rate constant k , measuring the passage of the supplement from the rumen to the abomasum, is obtained in a separate experiment in which the supplement is combined with a chromium marker which renders it completely indigestible. The effective percentage degradation, p , of the supplement, allowing for rate of passage, is shown to be p = a +[ bc/(c+k) ] (1- e -(e+k)t ) by time, t , after feeding. As t increases, this tends to the asymptotic value a+bc /( c+k ), which therefore provides an estimate of the degradability of the protein supplement under the specified feeding conditions. The method is illustrated by results obtained with soya-bean meal fed as a supplement to a dried-grass diet for sheep. The incubation measurements showed that 89% of the soya-bean protein disappeared within 24 h and indicated that it was all ultimately degradable with this diet. When the dried grass was given at a restricted level of feeding the allowance for time of retention in the rumen reduced the estimate of final degradability to 71% (69% within 24 h). With ad libitum feeding there was a faster rate of passage and the final degradability was estimated to be 66% (65% within 24 h).