The objective of this work was to compare pre-treatments with hydrogen peroxide at 1 and 2% (v/v) and pH at 11.5 used in elephant-grass (EC), mulatto-grass (HD364) lignocellulosic biomass, bark of rice (CA) and leaves of the pineapple (AB) plant to estimate the production of second-generation ethanol (2G). The experimental design was completely randomized in a 4x2 factorial scheme with four replicates per treatment. Enzymatic hydrolysis was performed with the pretreated biomasses using Cellic/CTec2 enzyme. 7 ml of the hydrolyzed material and 0.14 g of non-isolated Saccharomyces cerevisiae yeast were used. After this dilution the material was taken to a styrofoam oven and measured at the temperature and volume of the gas produced at the time of 0,04; 0,08; 0,2; 0,63; 0,92; 1,19; 1,93; 2,19; 2,82; 3,15; 3,94 hours post-incubation. The results showed that the observed differences between the pretreatments influenced the yield in 2G ethanol differently between the biomasses, being the CE that presented the best yield, however, when evaluating the yield by the peroxide content within each biomass, both AB and EC showed no differences between pre-treatments. HD364 had better yield with 1% peroxide and CA with 2% peroxide.