Abstract: In this study, we constructed recommendation fertilization index system using rice as experimental materials at Jiangdu City in Jiangsu Province based on field experiments from 2005 to 2011, and determined N doses using soil fertility balance method, and determined P, K doses using nutrients abundance and deficiency index method. The results showed that when the target yield for rice was 6825-9270 kg/ha, averaging 8145 kg/ha, fertilizer N for rice was 146.4-257.4 kg/ha with an average of 207.5 kg/ha. Data indicated that when the the soil available P content exceeded 21 mg/kg, application of fertilizer P was unnecessary, and when the the soil available P content was ranged from 17 to 21 mg/kg, application amount of P2O5 should be 40.5 kg/ha, and when the soil available P content was ranged from 10 to 17 mg/kg, P2O5 application was 48 kg/ha, and when the soil available P content was below 10 kg/ha, application of P2O5 was 64.5 kg/ha. Fertilizer K should not be applied when the soil available K content exceeded 140 mg/kg, and application K2O was 225 kg/ha when the soil available K content was from 115 to 140 mg/kg, and application of K2O was 40.5 kg/ha when the soil available K content was ranged from 60 to 115 mg/kg, and application of K2O was 82.5 kg/ha when the soil available K content was below 60 mg/kg. Fertilization index system was based on the field experiments and its availabality was testified in the serve for soil testing and formulated fertilization.