Abstract This study evaluated the effects of physically effective neutral detergent fiber (peNDF), and undigested neutral detergent fiber (uNDF) on dry matter intake (DMI), eating behavior, ruminal fermentation, and total tract permeability for finishing cattle. Simmental heifers (n=6; 667.8±28.4 kg) were used in a replicated 3×3 Latin square (21-d periods) balanced for carry-over effects. Barley grain-based diets were formulated to provide treatments with: low peNDF and low uNDF (95% barley and supplement, and 0% straw; 0FOR); low peNDF but high uNDF (85% barley and supplement, 10% pelleted straw; PEL); or high peNDF with high uNDF (85% barley and supplement, 10% chopped straw; ST). Treatment did not affect DMI (P=0.93). Heifers fed 0FOR spent less (P< 0.01) time eating vs. ST, where PEL was intermediate. Time spent ruminating was greater (P=0.01) for ST than 0FOR, with PEL not differing. Heifers fed 0FOR had fewer meals/d than ST (P< 0.01) and meal size was greater for 0FOR than PEL and ST (P< 0.01). Feeding 0FOR reduced (P=0.02) mean ruminal pH (5.65 vs. 6.07) and increased (522 vs. 288 min/d, P=0.02) the duration that pH was < 5.5 relative to ST with PEL being intermediate (5.82 and 504 min/d, respectively). Heifers fed 0FOR had greater (112.5 vs. 59.9 sec, P=0.05) time between contractions and lesser (61.45 vs. 94.58 mmHg′sec, P=0.05) contraction area compared with cattle fed ST, while PEL remained intermediate (76.73 sec and 73.29 mmHg×sec, respectively). Permeability of the gastrointestinal tract to Cr-EDTA and intestine to Co-EDTA did not differ (P≥0.11) among treatments. The data suggest that increasing uNDF improved ruminal function, but increasing uNDF with peNDF is needed to decrease the risk of ruminal acidosis in finishing cattle. The results further suggest that the use of uNDF and peNDF may provide useful information over NDF inclusion when evaluating fiber adequacy of finishing diets.