We used established long-term experimental P-amended plots in freshwater marshes of northern Belize to determine the impact of P input on nitrogen (N) fixation. Marshes with different conductivities and sulfate concentrations were selected to elucidate the effect of salinity and the contribution of sulfur reducing bacteria to the overall N fixation. Rates of N fixation in sediment, roots, and cyanobacterial mats was measured in laboratory incubation experiments (acetylene reduction assay calibrated by 15 N 2 reduction assay) with and without the addition of sodium molybdate (sulfur reducing bacteria inhibitor). P has increased macrophyte primary production significantly, which led to the rapid elimination of cyanobacterial mats and the elimination of autotrophic N fixation. P addition enhanced heterotrophic N fixation in both the sediments and rhizosphere due primarily to increased C supply to the sediment. When expressed on a dry weight basis, root associated N fixation was higher than sediment N fixation, but the contribution of the root associated fixation to the total N fixation was small when expressed per square meter. Sulfur reducing bacteria were an important component of N fixation, contributing from 20 to 53% to the overall N fixation. A simple N budget was created to determine if N demands are met following P addition. The heterotrophic N fixation substituted in part for autotrophic cyanobacterial N fixation when P limitation was alleviated.