Dinitrogen (N(2)) fixation is widely recognized as an important process in controlling ecosystem responses to global environmental change, both today and in the past; however, significant discrepancies exist between theory and observations of patterns of N(2) fixation across major sectors of the land biosphere. A question remains as to why symbiotic N(2)-fixing plants are more abundant in vast areas of the tropics than in many of the mature forests that seem to be nitrogen-limited in the temperate and boreal zones. Here we present a unifying framework for terrestrial N(2) fixation that can explain the geographic occurrence of N(2) fixers across diverse biomes and at the global scale. By examining trade-offs inherent in plant carbon, nitrogen and phosphorus capture, we find a clear advantage to symbiotic N(2) fixers in phosphorus-limited tropical savannas and lowland tropical forests. The ability of N(2) fixers to invest nitrogen into phosphorus acquisition seems vital to sustained N(2) fixation in phosphorus-limited tropical ecosystems. In contrast, modern-day temperatures seem to constrain N(2) fixation rates and N(2)-fixing species from mature forests in the high latitudes. We propose that an analysis that couples biogeochemical cycling and biophysical mechanisms is sufficient to explain the principal geographical patterns of symbiotic N(2) fixation on land, thus providing a basis for predicting the response of nutrient-limited ecosystems to climate change and increasing atmospheric CO(2).