The family Anacardiaceae has worldwide, mainly tropical, distribution and an extensive fossil record. However the phylogenetic relationships inside the family are not well known although some attempts to discover them using different types of characters have been made. Here, the leaf architecture is used to generate a phylogenetic hypothesis for the family. The analysis indicates that the five tribes proposed by Engler are not supported. Instead, two main sister lineages including most members of the family are the most distinctive feature of the cladogram. This same result has been found with other characters like the wood anatomy and molecules. The rest of the genera, with former Spondiadeae or Rhoeae affinities, are basal with reference to the two clades. This preliminary result suggests that the history of the family Anacardiaceae is ancient as members of the two main lineages have been found as fossils as early as the Eocene and probably the Cretaceous. A great diversity of anacardiaceous fossils from Mexico indicates that the family has a long and complex history in low latitude North America and that this could have been an important diversification area for some clades inside the family. Leaf architecture categorization was necessary as most anacardiaceous fossils are leaf impressions, making the study of morphological features and their evolution inside the family imperative to understand extant and extinct species.

Key words: Anacardiaceae, biogeography, leaf architecture, phylogeny, Tertiary