Our objective was to determine the long-term effect of feeding tannin extracts to lactating cows on performance and nitrogen (N) utilization efficiency. Thirty-four multiparous lactating cows (711 ± 59 kg BW; 113 ± 19 DIM) were randomly assigned to 3 dietary treatments in a randomized complete block design. Following a 2-wk covariate adjustment period, cows were fed their assigned treatment diets for 13 weeks. Rice hull was removed from a 54:46 forage to concentrate ratio (%DM) TMR as a tannin extract mixture from Quebracho and Chestnut trees (2:1 ratio) was included at 0 (control), 0.45, and 1.8% of dietary DM. Urinary excretion (g/d) was calculated as 0.0283 × BW (kg) × MUN (mg/dL). There was no interaction between dietary treatments and experimental week. Overall, milk yield (48.7 kg/d), fat-and-protein corrected milk (46.7 kg/d), milk fat content (3.89%) and yield (1.86 kg/d) and true protein yield (1.45 kg/d) were not affected by dietary treatments. There was a linear increase in DMI (29.2 to 30.9 kg/d), a linear decrease in milk/DMI (1.67 to 1.57 kg/kg) and a linear decrease in MUN (12.2 to 10.8 mg/dL) with incremental levels of tannin extracts in the diet (all P < 0.01). There was a quadratic effect (P < 0.01) for milk true protein content, (2.96, 3.13, and 3.00% for control, 0.45 and 1.8% tannin extract, respectively) and a tendency for linear (P = 0.07) and quadratic (P = 0.06) response for BW gain (0.31, 0.16 and 0.44 kg/d for control, 0.45 and 1.8% tannin, respectively). Intake of N increased linearly (782, 795, and 820 g/d) and N utilization efficiency (milk N/ intake N) decreased linearly (0.30, 0.30 and 0.28) for control, 0.45 and 1.8% tannin, respectively. An 11% reduction (P < 0.01) in estimated urinary N excretion was observed between control and 1.8% tannin extract. In this study, adding tannin to the diet reduced feed efficiency, but had a positive effect on milk protein content at 0.45% tannin extract level. Feeding tannins at 1.8% in the diet may reduce dietary protein degradation and urinary N excretion without affecting milk protein yield.