Rice is one of the main staple food crops and its demand is rapidly increasing in West Africa, hence there is need to increase its production at lower production costs through adequate N management with appropriate varieties.Nitrogen (N) management is one of the key inputs in rice production, especially in the savanna soils which is usually associated with widespread N deficiency. The focus of this research was to evaluate the performance of rice under different N fertilizer management options. Field experiments were conducted in 2012 and 2013 at the research farm of the Institute for Agricultural Research, Talata Mafara (irrigated) and Kadawa (rain-fed). The treatment combinations included two granular forms of N as Urea Super Granule (USG) and Granular Urea (GU) at different rates [0, 45.1, 72.2, 117.3 kg N/ha and a rate based on leaf colour chart (LCC)] with three varieties of rice (FARO 55, FARO 57 and FARO 52) as the test crops. The experimental design was Randomized Complete Block Design (RCBD) and replicated thrice. The second phase of the study involved the use of Quantitative Evaluation of the Fertility of Tropical Soils (QUEFTS) model to simulate rice performance as a decision support tool for site-specific fertilizer recommendation in the study areas. Crop growth attributes generally increased with increased N application for both USG and GU, but with delay in days to heading. Application of USG at 117.3 kg N/ha significantly (p<0.01) increased yield and yield attributes, except for grain-, phosphorus- and potassium- harvest indices where application of GU based on LCC gave the highest values. Interaction between N management and varieties indicated that application of USG at 117.3 kg N/ha on FARO 52 resulted in significant rice yield increase.Paddy yield was highly and positively correlated with total biomass (r= 0.883**), straw yield (0.733**), dry matter (0.351**), plant height (0.158**) and N uptake (0.701**); but weak and non-significant negative relationship existed between soil pH (- 0.021) and available P (- 0.055). The indigenous soil nutrients supplies (INS, IPS and IKS) were 25.26 - 94.50 kg N ha-1, 5.78 - 28.84 kg P ha-1 and 47.48 - 156.30 kg K ha-1with fertilizer recovery fractions of 0.25 - 0.61, 0.09 - 0.48 and 0.27 - 0.69 for N, P and K respectively at both locations. The recommended fertilizer dosage modeled by QUEFTS were 104 - 140 kg N ha-1, 69 - 94 kg P ha-1 and 23 - 46 kg K ha-1for dry season (Talata Mafara) and 92 - 130 kg N ha-1, 34 - 56 kg P ha-1 and 10 - 19 kg K ha-1 for wet season (Kadawa). It was higher than the conventional „blanket fertilizer‟ rate recommended over the years. The model also recommends higher N with lower K for FARO 55, but lower K for FARO 57 and FARO 52; thereby advocating for site specific nutrient management (SSNM) through the use of different fertilizer combinations for different soil conditions. Field validation of the QUEFTS model showed a good agreement between observed and simulated yields (R2 = 0.85, RSME = 0.93, RSMEn =< 30% and d-stat = 0.71), thus confirming better performance of QUEFTS in the rice ecosystem under different nitrogen regimes. Partial economic analysis of rice production at both locations revealed that application of USG at 117.3 kg N/ha with FARO 52 gave the highest gross margin of â‚Â�237,907.22k/ha and a profit of â‚Â�2.81k per naira invested. Results obtained from this study revealed that FARO 52 using USG had better paddy yields and highest NUE.Further studies are also required to establish the appropriate use of leaf colour chart in the study areas.