Every year, between 2.5 and 4.0 million km2 of vegetation are burned around the world, emitting an annual average of 2,013 Tg of C, at least 50% of which is estimated to represent net emissions that are not compensated for by the regeneration of vegetation. In Colombia, the official statistics of burned areas are not well known, but it is estimated that burned areas vary between 15,000 and 20,000 km2 per year. The main objective of this study was the development and application of a new algorithm for mapping burned areas in Colombia, using a synergistic combination of reflectance images and thermal anomalies detected by the MODIS sensors installed on NASA’s TERRA and AQUA satellites. Upon applying the developed algorithm, a burned surface area in Orinoquía of 998,473 ha was obtained for the month of February, 2007. The validation of the algorithm was performed using high spatial resolution Landsat images and the comparison of the data with the global MODIS MCD45A1 burned area product. The developed algorithm performed very similar to MCD45A1, with an overall accuracy of 79% in both cases. The new algorithm, which was developed for the mapping of burned areas in Colombia, can be used to complement deforestation and forest degradation monitoring procedures that are being implemented in Colombia in the context of the REDD+ mechanism initiation.