The digestibility of high-fibre diets and digesta passage was measured in growing pigs and attempts were made to predict the in vivo digestibility from in vitro data. In Experiment 1, six diets were formulated to measure digestibility using low-fibre cassava starch, fish meal and soya bean meal in combination with six locally available fibre sources. Four green plants: banana sheaths ( Musa paradisiaca), duckweed ( Lemna minor), sweet potato vines ( Ipomoea batatas) and water spinach ( Ipomoea aquatica) were used for the BS, DW, SPV and WS diets, respectively. In addition two by-products, copra meal (CM) and tofu residues (TR), were used for the CM and TR diets, respectively. Digesta retention of solids was measured using chromium mordanted Para grass ( Brachiaria mutica) fibre and Co–EDTA was used for the retention of liquid. In Experiment 2, the digestibility of diets based on the fibre sources: brewer’s grains (BG), copra meal, cassava residues (CR), whole ground rice (WGR), coarse rice bran (RB) and tofu residues was measured in growing crossbred Baxuyen pigs. These fibre sources replaced 30% of a basal diet consisting of mainly maize, rice bran, soya bean meal and fish meal. The diets from Experiments 1 and 2 were also used to measure in vitro neutral detergent fibre (NDF) degradation, using faecal inocula taken from pigs fed on the same diets. In Experiment 1, the coefficient of total tract apparent digestibility (CTTAD) organic matter (OM), crude protein (CP), and neutral detergent fibre values of all, except for the BS diet, were similar and ranged from 0.82 to 0.84, 0.65 to 0.76 and 0.69 to 0.72, respectively. CTTAD for OM, CP and NDF of the BS diet were, however, significantly lower and 0.72, 0.60 and 0.32, respectively. Solids and liquid mean retention times (MRT) for the BS diet were 22.2 and 22.6 h and for the other diets, MRT ranges were 32.9–38.9 and 29.5–36.9 h, respectively, and were similar for the two markers ( P=0.31). In Experiment 2, CTTAD for OM of the WGR and RB diets were 0.73–0.77 and for NDF 0.57–0.64. These data were lower than those for the other diets ( P=0.01), which were similar and were 0.82–0.90 and 0.70–0.79, respectively. Crude protein CTTAD for the BAS diet (0.87) was significantly higher than that of the TR (0.73) and WGR (0.75) diets ( P<0.01). There was a high correlation between the NDF digestibility in vitro at 36 h and in vivo in Experiment 1 ( r 2=0.92, R.S.D.=0.035). The results indicate that the green plants and high-fibre by-products had relatively high CTTAD of OM, CP and NDF and could be reasonable feed resources for pigs, even though intake will restrict the levels of inclusion from 200 to 350 g kg −1 of the diet. However, lignin and silica limited the utilisation of coarse rice bran, and a short retention time and the high-fibre and tannin contents reduced the digestibility of the banana sheaths. In vitro NDF degradation seemed to be a reliable predictor of in vivo NDF digestibility.