Abstract This study aimed to evaluate the inclusion of tannins extract (TN) associated to two levels of urea (UR) in the supplement of grazing Nellore steers on ruminal microbial richness and diversity estimators. The study used eight Nellore steers (293 kg ± 5.6 BW), rumen cannulated grazing Urochloa brizantha cv. Marandu distributed in a double Latin square 4 x 4 (four experimental periods and four supplements). Steers were supplemented at 3 g/kg BW with soybean meal, corn, mineral mix and two levels of UR (3% or 5%) associated or not to TN (SilvaFeed-ByPro at 0.7 g / kg). On d 26 of each experimental period, samples of ruminal content were collected after supplementation (+3h), V3/V4 regions of 16SrRNA gene was sequencing using the Illumina MiSeq. Quantitative Insights into Microbial Ecology (QIIME v.1.9.1) were used to filter reads and determine Operational Taxonomic Units (OTUs) and Alpha_diversity.py in QIIME was used to calculate ACE, Chao1, Shannon, and Simpson indices. Richness and diversity estimators were compared between UR level and TN addition using a Friedman test and a Dunn’s post-hoc test using R. Sequencing produced 709,456 reads from16 samples. After trimming, the median number of reads was 81,005 per sample with good’s coverage median of 0.98. There was no effect of UR levels on the indexes of richness ACE (2487.4), Chao1 (2497.3), diversity Shannon (8.95) and Simpson (0.990) (P > 0.05). TN addition in the supplement increased ACE (2478.7 to 2483.1) and Chao 1 (2495.3 to 2507.9), had no effect in the diversity Shannon (8.93) and reduced Simpson index (0.993 to 0.990) (P < 0.05), suggesting higher microbiome growing and a more efficient rumen due to lower numbers of metabolic pathways. Higher richness and lower diversity indexes suggest that TN may improve the rumen microbiome growing and efficiency in grazing steers.