An experimental investigation in a 100 kW(th) bubbling fluidized bed gasifier (FBG) was conducted to study the effect of primary measures on the tar composition in the gas. The measures tested included the variation of the stoichiometric ratio, SR, defined as the ratio of fed to stoichiometric air, secondary-air injection in the freeboard, and addition of lime and alumina to the reference bed material (ofite). The reactor was operated nearly adiabatic, with the aim of simulating the performance of industrial autothermal air-blown FBG units. An optimum value of SR was found at approximately 0.3, where the cold gasification efficiency was maximum (70%) and the tar concentration was reduced by a factor of 3 with respect to reference conditions. Addition of secondary air (keeping SR constant) reduced the gravimetric tar and the water-soluble tar compounds up to 20 and 30%, respectively; a fixed ratio of secondary to total air of 11% was investigated in this work for three valued of SR. The use of lime allowed for the reduction of the overall tar content up to 50%, whereas the decrease using alumina was rather low. None of the measures tested effectively reduced the heavy tar compounds, so the minimum value computed for the tar dew point was 170 degrees C.