Rates of direct denitrification were measured using ¹⁵N isotope tchniques in intact sediment cores from fringe and basin mangroves in Terminos Lagoon, Mexico. Sediments were injected with ¹⁵NO₃^- and the distribution of ¹⁵N was measured over time in the head space, overlying water, and sediments. Experiments included an investigation of spatial variation in denitrification rates by comparing results from the fringe and basin mangroves in the rainy season (July 1991); a second experiment was to determine the effect of 3 different NO₃^- concentrations (25, 100, 200 umol core^-115N-KNO₃^-) on denitrification rates and was performed in the fringe mangrove during the 'Norte' season (January 1992). Highest ¹⁵N-N₂ fluxes were measured in the fringe mangrove at 9.4 umol m^-2 h^-1, while denitrification rates in the basin mangrove ranged from 1.9 to 4.5 umol m^-2 h^-1. ¹⁵N-N₂ fluxes in sediment cores from the fringe mangrove were significantly higher (4.5 to 7.7 umol m^-2 h^-1) in cores enriched with 200 umol core^-115N-KNO₃^- compared to cores enriched with 25 and 100 umol core^-115N-KNO₃^- (<1 umol m^-2 h^-1). Most of the applied ¹⁵N was recovered as particulate nitrogen in the sediment and a small fraction reduced to NH₄⁺ in both experiments. The low denitrification rates observed in the fringe and basin mangroves indicate that the capacity for sediment denitrification is limited by low NO₃^- availability. Previous nutrient exchange studies concluded that the fringe mangrove was a 'sink' of NO₃^-since sediment uptake of NO₃^- was assumed lost through denitrification. Results from this study show <10% of sediment NO₃^- uptake in fringe mangroves may be lost to denitrification; the remainder being immobilized in the sediment.