BackgroundRutaceae subfamily Rutoideae (46 genera, c. 660 species) is diverse in both rainforests and sclerophyll vegetation of Australasia. Australia and New Caledonia are centres of endemism with a number of genera and species distributed disjunctly between the two regions. Our aim was to generate a high-level molecular phylogeny for the Australasian Rutoideae and identify major clades as a framework for assessing morphological and biogeographic patterns and taxonomy.Methodology/Principal FindingsPhylogenetic analyses were based on chloroplast genes, rbcL and atpB, for 108 samples (78 new here), including 38 of 46 Australasian genera. Results were integrated with those from other molecular studies to produce a supertree for Rutaceae worldwide, including 115 of 154 genera. Australasian clades are poorly matched with existing tribal classifications, and genera Philotheca and Boronia are not monophyletic. Major sclerophyll lineages in Australia belong to two separate clades, each with an early divergence between rainforest and sclerophyll taxa. Dehiscent fruits with seeds ejected at maturity (often associated with myrmecochory) are inferred as ancestral; derived states include woody capsules with winged seeds, samaras, fleshy drupes, and retention and display of seeds in dehisced fruits (the last two states adaptations to bird dispersal, with multiple origins among rainforest genera). Patterns of relationship and levels of sequence divergence in some taxa, mostly species, with bird-dispersed (Acronychia, Sarcomelicope, Halfordia and Melicope) or winged (Flindersia) seeds are consistent with recent long-distance dispersal between Australia and New Caledonia. Other deeper Australian/New Caledonian divergences, some involving ant-dispersed taxa (e.g., Neoschmidia), suggest older vicariance.Conclusions/SignificanceThis comprehensive molecular phylogeny of the Australasian Rutoideae gives a broad overview of the group’s evolutionary and biogeographic history. Deficiencies of infrafamilial classifications of Rutoideae have long been recognised, and our results provide a basis for taxonomic revision and a necessary framework for more focused studies of genera and species.