Abstract In two experiments, four rumen-fistulated steers were used in a Latin square design arrangement with each rotation period of 15 days. Yucca schidigera extract (YE) was added at the rate of 250 mg kg −1 of mixed feed. In the first experiment, four isonitrogenous high-roughage (HR) feeds were tested: HR + soybean meal (R1); R1 + YE (R2); HR + urea (R3); R3 + YE (R4). Feed intakes (FI) (kg day −1) were 11.4, 11.2, 11.3, and 11.0 for R1, R2, R3 and R4, respectively. Ruminal pH and rumen ammonia-N (RAN, mg dl −1) for R1, R2, R3 and R4 (pooled values for 0, 3, 6, 9 h postfeeding, HPF) were: 6.5, 11.5; 6.6, 10.2; 6.6, 15.1; 6.6, 12.9. Plasma ammonia-N (μg ml −1) and plasma urea-N (mg dl −1) for R1, R2, R3 and R4, respectively, were: 1.13, 13.64; 1.24, 14.79; 1.19, 16.11; and 1.04, 14.7 In Trial 2, the four isonitrogenous, high-concentrate (HC) diets were: HC (barley) + soybean meal (C1); C1 + YE (C2); HC + urea (C3); C3 + YE (C4). Average daily FI (kg) were: 14.2, 13.5, 14.2 and 13.7 for C1, C2, C3 and C4, respectively. Ruminal pH and RAN for C1, C2, C3 and C4, respectively, were: 5.81, 7.92; 5.82, 6.88; 6.09, 10.85; 6.00, 10.50 The plasma ammonia-N (PAN) and plasma urea-N (PUN) for C1, C2, C3 and C4 were: 1.17, 14.00; 0.89, 11.52; 1.15, 13.87; 1.15, 13.76. In the HR trial, rumen fluid concentrations of different volatile fatty acids (VFA) (averaged across time) were not different among treatments, except that isovalerate concentration in R1 was higher ( P < 0.05) than R2. Numerically, acetate concentration was lower in R2 vs. R1, and propionate, isobutyrate and valerate concentrations were higher in R4 vs. R3. There was less fluctuation in VFA concentrations at different HPF in R4 vs. R3. Total VFA concentration was higher in R1 vs. R2 and R4 vs. R3. The acetate:propionate ratio was lower in R1 vs. R2 and R4 vs. R3. In the HC trial, acetate, propionate, isobutyrate and valerate concentrations were lower in the YE diets. Total VFA concentrations and acetate:propionate ratios were generally lower in the YE diets. The results showed a tendency for YE to reduce RAN, PAN and PUN, which could lead to more efficient utilization of dietary nitrogen, particularly when HR diets supplemented with urea are fed.