Species radiations provide unique insights into evolutionary processes underlying species diversification and patterns of biodiversity. To compare plant diversification over a similar time period to the recent cichlid fish radiations, which are an order of magnitude faster than documented bird, arthropod, and plant radiations, we focus on the high-altitude flora of the Andes, which is the most species-rich of any tropical mountains. Because of the recent uplift of the northern Andes, the upland environments where much of this rich endemic flora is found have been available for colonization only since the late Pliocene or Pleistocene, 2-4 million years (Myr) ago. Using DNA sequence data we identify a monophyletic group within the genus Lupinus representing 81 species endemic to the Andes. The age of this clade is estimated to be 1.18 -1.76 Myr, implying a diversification rate of 2.49 -3.72 species per Myr. This exceeds previous estimates for plants, providing the most spectacular example of explosive plant species diversification documented to date. Furthermore, it suggests that the high cichlid diversification rates are not unique. Lack of key innovations associated with the Andean Lupinus clade suggests that diversification was driven by ecological opportunities afforded by the emergence of island-like habitats after Andean uplift. Data from other genera indicate that lupines are one of a set of similarly rapid Andean plant radiations, continental in scale and island-like in stimulus, suggesting that the high-elevation Andean flora provides a system that rivals other groups, including cichlids, for understanding rapid species diversification.leguminosae ͉ Lupinus ͉ phylogeny ͉ species diversification S tudies of rapid episodes of species diversification or species radiations have provided a continuously rich source of new insights into the evolutionary processes underlying diversification and modern patterns of biodiversity for the last 150 years (1-6). A striking feature of species radiations is the discrepancy between the very high rates of species diversification reported for cichlid fish radiations in east African rift lakes and all bird, arthropod and plant radiations (2,7,8). This discrepancy has been attributed in part to the recency of the fish radiations (7). Accurate measurement of peak episodes of diversification embedded within older radiations requires fully sampled and resolved phylogenies (9). In the absence of such phylogenies, measurements of diversification rates for most radiations are less precise because they average out episodes of faster and slower speciation. Furthermore, these approaches measure net diversification rates, and the effects of extinction are not assessed. If speciation is concentrated in the early phases of radiations (3,7,10,11), this could explain the discrepancy between the exceptional rates of species diversification reported for very recent [Ͻ2 million years (Myr)] lacustrine fish and other documented bird, arthropod, or plant radiations, which are generally older (Ͼ5...