The mobile bag technique (MBT) was used in two ileorectostomized and two intact pigs with duodenal cannulae, in order to estimate the precaecal and total crude protein digestibility and amino acid (lysine, methionine, cystine, threonine, tryptophan) absorption of ten single feedstuffs (field bean, lupin, rapeseed meal, fish meal, barley, wheat) and three diets. For each feed 36 bags were applied per animal. After a peptic predigestion in vitro, application of the bags through the duodenal cannula into the intestine and their passage through the gut, the residues remaining in the bags were pooled to collective samples and just as the original feed material analysed on dry matter, nitrogen and amino acids. The directly determined MBT-digestibilities (BV) and the values obtained by regressive calculations (BVb) have been compared with apparent digestibilities (sV) conducted from conventional digestion experiments. 74 per cent of all BV data was higher than the corresponding sV values. Similarities of BV with sV showed in 58 per cent, those of BVb with sV only in 17 per cent. In all cases differences of > +/- 4 digestibility units showed. Checking additivity of diet-BV by means of calculations from the BV of their single ingredients did not generally lead to the same results as the directly estimated diet-BV; in some cases the results were contrary. Therefore the use of MBT for mixed feed diets has to be further investigated. For single feedstuffs the MBT can be used successfully as a rapid and simple technique when regressive calculations are employed.