The experiment was conducted to compare the mobile nylon bag technique (MNBT) with conventional digestibility techniques for the estimation of digestible energy (DE) in pig diets. Factors investigated were the effect of washing after retrieval, predigestion time, particle size, retention time of the bags, and the effect of the basal diet. Retention time and predigestion time had no effect on DE. DE values obtained with unwashed bags (13,57 MJ/kg) were highly significantly (P ≤ 0,01) lower than either DE values obtained with washed bags (15,01 MJ/kg) or values obtained with the conventional technique (14,87 MJ/kg). Although particle size appeared to play an important role, DE values obtained with washed bags, for the different particle sizes, differed non-significantly from the reference DE value (14,87 MJ/kg). DE values from material with the 3-mm particle size (14,52 MJ/kg) were significantly (P ≤ 0.01) lower and DE values, with an 800–µm particle size (15,49 MJ /kg), significantly (P ≤ 0,05) higher, than the DE value obtained with the 1-mm particles (15,02 MJ/kg). The 1-mm particles correlated best with the reference DE value and differed from it by only 1,0%. Washed bags, with or without predigestion, which were milled through a 1-mm sieve size correlated best with the reference DE value. According to an analysis of correlation between DE versus particle size using the MNBT method, a mean diameter of 1,16 mm should give the same predicted DE value as that obtained by the conventional technique. The basal diet appeared to influence the DE value obtained with the MNBT, in that a high crude fibre level (8%) in the diet was associated with a lower DE value (14,53 MJ /kg) than that (15,17 MJ/kg) obtained with a low fibre (2%) diet, although these differences were not significant. With high tannin feed samples in the bags, the DE value obtained with a low-tannin basal diet (14,20 MJ /kg) appeared to be higher than the DE value obtained with a high-tannin basal diet (13,81 MJ /kg).