In the lowland Maya area, pollen records provide important insights into the impact of past human populations and climate change on tropical ecosystems. Despite a long history of regional paleoecological research, few studies have characterized the palynological signatures of lowland ecosystems, a fact which lowers confidence in ecological inferences made from palynological data. We sought to verify whether we could use pollen spectra to reliably distinguish modern ecosystem types in the Maya lowlands of Central America. We collected 23 soil and sediment samples from eight ecosystem types, including upland, riparian, secondary, and swamp (bajo) forests; pine savanna; and three distinct wetland communities. We analyzed pollen spectra with non-metric multidimensional scaling (NMDS), and found significant compositional differences in ecosystem types' pollen spectra. Forested sites had spectra dominated by Moraceae/Urticaceae pollen, while non-forested sites had significant portions of Poaceae, Asteraceae, and Amaranthaceae pollen. Upland, bajo, and riparian forest differed in representation of Cyperaceae, Bactris-type, and Combretaceae/Melastomataceae pollen. High percentages of pine (Pinus), oak (Quercus), and the presence of Byrsonima characterized pine savanna. Despite its limited sample size, this study provides one of the first statistical analyses of modern pollen rain in the Maya lowlands. Our results show that pollen assemblages can accurately reflect differences between ecosystem types, which may help refine interpretations of pollen records from the Maya area.