Neotropical forests are the most diverse terrestrial ecosystems on Earth. Understanding the earliest phases in the origin of Neotropical forests and how they were affected by changing geography and climates, beginning in the Early Cretaceous (~145 Ma), is crucial for understanding the initial assembly of these hyper-diverse plant communities. Unfortunately, Cretaceous floras from northern South America are very poorly known, making it difficult to develop secure hypotheses concerning the origin of angiosperm-dominated ecosystems in this area. The aim of my work was to expand the fossil record of Neotropical Cretaceous floras by exploring fossiliferous localities and paleobotany collections from Colombia. This ongoing project is made in conjunction with a scientific team from the Center for Tropical Paleoecology and Archaeology of Smithsonian Tropical Research Institute (CTPA-STRI). I explored kaolinite deposits, in Boyacá and Santander Departments, from the Early Cretaceous Cumbre Formation (Valanginian, ~140.2-136.4 Ma). To date there were not reports of fossil plants from the Cumbre Formation. Preliminary analysis from the Cumbre sediments indicates a flora dominated by typical Lower Cretaceous palynomorphs (i.e. Classopollis sp.). Charcoalification and very high thermal maturity of the organic fraction suggests a depositional environment exposed to fire and/or high temperatures. As part of this exploration, I also visited the paleobotanical collections at CTPA-STRI in Panama, where leaf and fruits macrofossils from the Maastrichtian (Upper Cretaceous, ~70 Ma) Guaduas Formation (Boyacá) are deposited. The temporal framework of the Guaduas formation, only ~5 million years before the Cretaceous-Paleogene boundary (K/P), makes the fossil material found a critical source to address questions on how tropical floras suffered a major change from non-angiosperm dominated floras to angiosperm-dominated ones at the end of the Cretaceous. My work, in collaboration with the CTPA-STRI, team was focused on estimating quantitatively the proportion of plant extinction during the K/P in Neotropical forests by quantifying the number of leaf and fruit morphotypes present in the Guaduas flora, and comparing them with two Paleocene (~57 Ma, around 5 Ma after the K/P) floras from Colombia (Cerrejón and Bogotá Formations). Our estimates suggest a rate of extinction of about 80%, which is concordant with previous estimations from pollen data. Moreover, the Guaduas florasuggests that the Late Cretaceous of Colombia lacks the most abundant families present in Paleocene fossil assemblages as well as the families most characteristic of modern Neotropical rainforests. This evidence is consistent with a major floral turnover at the end of the Cretaceous. These approaches are contributing to our understanding of the mechanisms that produced the high plant diversity of the extant Neotropical forests, and to a deeper understanding of their phylogenetic history and biogeography as well as the potential connections between evolutionary patterns and climatic perturbations in the Cretaceous globally warm world. Comparing vegetational patterns before and after the end-Cretaceous major extinction event may also be relevant to understanding the present day climate change/diversity crisis scenarios.