Tropical legumes like sunn hemp (Crotolaria juncea L.) and aeschynomene (Aeschynomene evenia L.) have potential as alternative cover crops in tropical regions. The objective of this study was to evaluate the N mineralization rates of three cover crops [aeschynomene (AE), sorghum sudangrass (Sorghum sudanense L.), and sunn hemp (SH)] residues used to amend a calcareous gravelly soil in order to predict nitrogen (N) availability for uptake by subsequent crops. The soil amended with cover crop residues was incubated in columns under ambient and controlled temperatures. Mineralization rates were determined with a leaching/incubation method. The mineral N ratios between ambient and controlled temperature ranged from 1.39 to 2.8 with an average of 2.1. The measured cumulative mineral-N indicated that legumes were far more effective than non-leguminous cover crops in increasing N availability in the soil. The maximum values of mineral N recovered were 19, 7, and 8% of added N for AE, SH, and sorghum sudan (SS), respectively, when incubated at ambient temperature. Based on the Mean Square Error (MSE) for the equations describing mineralization kinetics, the power function was better than other models for all treatments, except SS at both the temperatures. The comparison of goodness of fit between the observed and predicted values indicated that the power function and parabolic equation produced generally good predictions for AE, SH, and control treatments at ambient temperature. However, at controlled temperature, the power function and the zero-order equations gave the best predictions for all treatments.