RATIONALE It is important for the enforcement of the CITES treaty to determine whether agarwood (a resinous wood produced in Aquilaria and Gyrinops species) seen in trade is from a plantation that was cultivated for sustainable production or was harvested from natural forests which is usually done illegally. METHODS We analyzed wood directly using Direct Analysis in Real Time (DART™) ionization coupled with Time-of-Flight Mass Spectrometry (TOFMS). Agarwood was obtained from five countries, and the collection contained over 150 samples. The spectra contained ions from agarwood-specific 5,6,7,8-tetrahydro- 2-(2-phenylethyl)chromones as well as many other ions. The data was analyzed using either kernel discriminant analysis or kernel principal component analysis. Probability estimates of origin (wild vs cultivated) were assigned to unknown agarwood samples. RESULTS Analysis of the DART-TOFMS data shows that many of the chromones found in cultivated and wild agarwood samples are similar; however, there is a significant difference in particular chromones that can be used for differentiation. In certain instances, the analysis of these chromones also allows inferences to be made as to the country of origin. Mass Mountaineer™ software provides an estimate of the accuracy of the discriminate model, and an unknown sample can be classified as cultivated or wild. Eleven of the thirteen validation samples (85%) were correctly assigned to either cultivated or wild harvested for their respective geographic provenance. The accuracy of each classification can be estimated by probabilities based on Z scores. CONCLUSIONS The direct analysis of wood for the diagnostic chromones using DART-TOFMS followed by discriminant analysis is sufficiently robust to differentiate wild from cultivated agarwood and provides strong inference for the origin of the agarwood.