Monitoring movement of nitrate nitrogen (NO 3 -N) from agricultural fields to subsurface and groundwater resources have received considerable interest worldwide. Agricultural activities are the most important source of NO 3 -N groundwater pollution. Also, groundwater is largest source of drinking water in the world. This research was conducted in a farm in Marvdasht, Fars province. The study site was selected to have three source of irrigation water includings groundwater, rivers and wastewater. In each treatment seven sampling locations with different irrigation water source, were randomly selected and soil samples were collected from 5 layers of soil profile (0-30, 30-60, 60-90, 90-120,120-150). After preparing the samples, the NO 3 -N concentration and soil chemical and physical properties were determined. The results showed the soils irrigated with wastewater, caused the maximum (12.27 mg / kg) accumulation of NO 3 -N in the 90-120 cm soil layer and 54% of the NO 3 -N was below the alfalfa roots. In river and groundwater irrigated fields NO 3 -N accumulation in soil profile was not observed. Soils irrigated with wastewater and groundwater are similar in soil physical properties. No accumulation of NO 3 -N was observed in soils irrigated with groundwater due to lack of source emissions and soils which are irrigated with river are due to severe leaching. Nitrate nitrogen content of soils irrigated with river water and wastewater compared to groundwater were 56 and 54 percent. Soils irrigated with river water and wastewater with 97 and 67 kg per hectare nitrogen--nitrate levels in excess of plant needs a positive potential for NO 3 -N leaching. The results of the application of artificial neural networks - genetic algorithms and input sensitivity analysis features showed that soil physical properties are the most important parameters effecting leaching of NO 3 -N. Bulk density, degree of contraction and soil depth accounted for 36, 16 and 16% of NO 3 -N leaching in the soil profile. Keywords: Soil nitrate nitrogen, Net irrigation requirement, nitrate loss, Genetic Algorithm, Artificial Neural Networks, Sensitivity analysis
