The aim of this study was to evaluate the bacterial community composition by denaturing gradient gel eletrophoresis (PCR-DGGE). Eighteen steers Holstein × Gyr with average body 8 155 ± 5 kg d⁻¹ were randomly distributed in a completely randomized design with 3 treatments and 6 repetitions. The diet was calculated for average daily gain of 1.2 kg/d and an average weight of 240 kg, using the requirements for crossbred steers estimated by BR-Corte. Forage:concentrate ratio (based on DM) used was 60:40 and the animals received dietary treatment 1.2% DM of BW; 1.9% DM of BW and ad libitum intake, as maintenance, intermediate and high gain treatment, respectively. The daily feed intake was recorded and animals were weighed weekly each 28 d. The production of methane enteric by the animals was measured by open-circuit respiration chambers for 2 consecutive 24-h days. To assess the genetic diversity of the ruminal microbial community, 50 mL of rumen fluid samples were collected at the slaughter. DNA was extracted and processed by phenol-chloroform and bead beating method. PCR reaction used universal primers to amplify the V3 region to amplify the 16S rRNA of archaea Nested-PCR was performed to amplify a shorter region of the archaea 16S rRNA, using the primers ARC344f-GC/517r for archaea. PCR-DGGE patterns were analyzed using BioNumerics software 5.1 with which hierarchical cluster comparisons were carried out to group similar profiles and to generate a binary matrix of band classes. All the images were normalized using the internal control samples and the comparison among whole profiles was performed using the Dice similarity coefficient. The total number of the detected bands represented the species richness. Shannon-Wiener index was calculated based on relative band intensity and the total of number bands of each DGGE profile. The statistical analyses were done using the software R. The methane emission was affected by the treatments (P < 0.05) but there was no effect of the treatments on the richness index also by Shannon-Wiener (P > 0.05). No differences in archaeal population were detected between treatments.