Five feeds and five castrates were used to study the effects of washing rather than scrapping feces fr om bags on nutrient digestibility obtained with the mobile nylon bag technique (MNBT). Using unwashed bags, the MNBT and the conventional indicator method (CON) produced similar dry matter digestibility (DMD) with barley being the only ingredient where a difference (P < 0.05) was obtained between the two techniques. The linear regression equation was CON DMD = 5.39 + 0.92 MNBT DMD (r2 = 0.95). The MNBT overestimated DMD with a bias of less than 0.1%. When washed bags were used, differences (P < 0.05) were obtained between the two techniques for barley, corn and oats and the MNBT overestimated DMD with a bias of 6.6%. There were no differences in gross energy digestibility (GED) when obtained with the indicator method or the MNBT using unwashed bags. The regression line equation was CON GED = 3.85 + 0.94 MNBT GED (r2 = 0.95) and the MNBT overestimated GED with a bias of 1.4%. When washed bags were used, significant (P < 0.05) differences were obtained between techniques for GED with barley, corn, and oats. The MNBT overestimated GED with a bias of 7.3%. With unwashed bags, there were no differences in crude protein digestibility (CPD) obtained with the indicator method or the MNBT, with the exception of the mixed diet. The regression line equation was CON CPD = 28.16 + 0.60 MNBT CPD (r2 = 0.53) and the MNBT overestimated CPD with a bias of 3%. When washed bags were used, significant (P < 0.05) differences were obtained between techniques for barley, corn, oats, wheat and the mixed diet, and the MNBT overestimated CPD with a bias of 12.8%. The overall results of this experiment indicate that washing bags greatly overestimates nutrient digestibility coefficients obtained with the MNBT. Key words: Nylon bag, pigs, digestibility, protein, energy, cannula