The objective of this study was to evaluated fermentation losses and silage quality with the addition of different by-products in TMR. A completely randomized design was used with four treatments: Control (CTL, with hay, corn silage, cracked corn, soybean meal, wheat bran and minerals); CTL plus the addition of fresh orange pulp (FOP); CTL plus soybean hulls (SBH) or CTL plus wet brewery wastes (WBW), with six replicates. The ingredients were mixed and ensiled in 24 PVC experimental silos. After 56 days of ensiling, fermentation losses (effluents, dry matter and gases), and the TMR silage pH and chemical composition were analyzed; aerobic stability was evaluated over seven days after the silos were opened. Lower dry matter (P ? 0.0001) was observed in TMR ensiled with wet (WBW and FOP) compared to dry (SBH) by-products. Greater ADF (P = 0.031) was observed in SBH when compared to others by-products. WBW and FOP increased dry matter (P ? 0.0001) and effluent losses (P ? 0.0001) when compared to CTL and SBH. SBH treatment remained stable after 176-hour, which differ (P = 0.0015) compared with the other treatments (range from 94.5 to 99 hours of exposure to air). The density (kg DM/m3) changed (P ? 0.0001) among treatments. The inclusion of SBH in TMR silages had the greatest density, followed by CTL, WBW and FOP. The density was negatively correlated with DM losses ( r= - 0,81; P ? 0.0001). The pH differ (P = 0.003) among treatments. Highest pH of TMR silage was observed for SBH (3,67) and the lowest (3,56) for FOP. Soybean hulls successful improved aerobic stability and density, whereas aerobic stability and density were decreased and fermentation losses increased when fresh orange pulp and wet brewery were used. Little effects on chemical composition were observed, thereby agro-industrial by-products can be used in TMR silages as long as there is adequate product availability and pricing.