The present paper addresses NOx emissions from wood stoves through a computational fluid dynamics (CFD) modeling approach. The most significant route for NOx formation in traditional biomass combustion applications is the fuel NOx mechanism. The formation of fuel NOx is very complex and sensitive to fuel composition and combustion conditions. Thus, accurate predictions of fuel NOx formation in wood stoves, which constitute a wide range of compositions and states, rely heavily on the use of chemical kinetics with sufficient level of details. In the present work we use CFD together with three gas phase reaction mechanisms; one detailed mechanism consisting of 81 species and 1401 reactions, and two skeletal mechanisms with 49 and 36 species respectively to predict NOx emissions from wood stoves. The results show that; using the detailed mechanism as reference the 49-specie reaction mechanism predict similar results whilst the 36-specie mechanism overpredicts the total amount of fixed nitrogen emissions (NO, NO2, N2O, HCN, NH3) whilst underpredicting NOx emissions. Furthermore, the results indicate that even in these small-scale applications, air staging can be used to reduce the NOx emissions.