Oxy-fuel combustion is one of the existing alternatives towards CO2 capture in power plants and heat-intensive industries. Oxy-fuel combustion of coal has been extensively studied so far, from a lot of lab-scale facilities to some demo-scale plants. Nevertheless, oxy-fuel combustion of biomass has not received much attention and could widen the development of the technology. The combination of a CO2-neutral biofuel along with a capture and storage system, the so-called “Bio-CCS” concept, would lead to negative CO2 emissions technologies. In this paper, an experimental investigation concerning oxy-fuel combustion of torrefied biomass is presented and discussed. Two types of torrefied biomass are selected, one obtained from forestry residues (pine wood) and the other from agricultural wastes (blend of vineyard and corn stover residues). Experiments are conducted in a lab-scale entrained flow reactor, designed to fire pulverized solid fuels under O2/N2, O2/CO2 and O2/CO2/H2O atmospheres. The objective is to analyze the influence of replacing a fossil fuel (coal) by biofuels (torrefied biomass), paying attention to the burnout rate, the combustion efficiency and the NOx formation rates. Besides the effect of the biomass type, other operating variables are also accounted for during the discussion: O2 and H2O concentrations in the firing atmosphere. Recently, CO2 replacement by H2O has been highlighted as an improved oxy-combustion approach, so four different steam contents are here investigated: 0% (dry oxy-combustion), 10% (wet recycle oxy-combustion), 25% and 40% (towards oxy-steam combustion). From the experimental tests and results, the best O2/CO2/H2O combination leading to the maximum fuel conversion and the minimum NOx formation is obtained and justified.