This is a guide for the collection of reference samples of trees to enable the identification of species and/or geographical origin of woody material. It is an update of the sampling section of the GTTN standards and guidelines (Ekué 2014) and builds further on a discussion initiated during a workshop held in Hamburg at the Thünen Institute for Wood Research in 2014. If you are looking for support on how to collect test samples, see the UNODC guide (UNODC 2016). To enable the implementation of the different laws regulating the trade in illegal wood, reference databases for various timber tracking tools are urgently needed for at least the most traded and endangered tree species. The Global Timber Tracking Network (GTTN) is building a central database where not only the reference data can be stored but which will also function as a sample locator. Having a common sampling guide will facilitate meaningful exchange of samples. In addition, to optimise the use of wood/wood product identification (taxonomic identity or geographic origin) in support of law enforcement, the guide anticipates upcoming developments to combine (Paredes Villanueva 2018) different timber identification methods (Dormontt et al. 2015, Lowe et al. 2016) such as wood anatomy (Koch and Schmitt 2015, Helmling et al. 2018), DNA-based methods (Jolivet and Degen 2012, Blanc-Jolivet et al. 2018, Chaves et al. 2018), stable isotopes (Paredes-Villanueva et al. in preparation, Vlam et al. 2018), DART TOFMS (Lancaster and Espinoza 2012, Espinoza et al. 2015, Deklerck et al. 2017, Paredes-Villanueva et al. 2018) and NIRS (Pastore et al. 2011, Bergo et al. 2016, Snel et al. 2018). This sampling guide is written to make sharing of samples between researchers specialised in different timber tracking methods possible, as samples should ideally come from the same location in the tree, from the same individual and from well-identified trees when combining methods. This guide is intended for scientists, to provide all the information needed to get the most out of sampling campaigns for timber identification purposes. This information should allow setting up a sampling protocol adapted to the specific goal of the research project, the conditions of the sampling area and the background of the people who will do the sampling. Note that this guide is to collect reference samples and hence relatively high amounts of samples from different individuals are needed to take the variability of a species into account. Once reference data have been developed for a tree species for one or more identification methods, however, only one sample of an unidentified wooden object is often sufficient to determine its identity.