Residential wood combustion emissions are one of the major global sources of particulate and gaseous organic pollutants. However, the detailed chemical compositions of these emissions are poorly characterized due to their highly complex molecular compositions, non-ideal combustion conditions and sample preparation steps. In this study, the particulate organic emissions from a masonry heater using three types of wood logs, namely, beech, birch and spruce, were chemically characterized using thermal desorption in situ derivatization coupled to a GCxGC-ToF/MS system. Untargeted data analyses were performed using the comprehensive measurements. Univariate and multivariate chemometric tools, such as analysis of variance (ANOVA), principal component analysis (PCA) and ANOVA simultaneous component analysis (ASCA), were used to reduce the data to highly significant and wood type-specific features. This study reveals substances not previously considered in the literature as meaningful markers for differentiation among wood types.