To address the relationships between the amount of nitrogen fertilizer application and the yield of double cropping rice systems, we investigated the effects of a cultivation pattern of strong seedlings with increased planting density and reduced nitrogen application (SDN) on the morphological and physiological characteristics of double cropping rice. Our results indicated that the effects of SDN on the morphological characteristics of the single plant roots of double cropping rice were not significant, but the morphological characteristics of the population roots were largely different. Specifically, SDN significantly increased the morphological indexes of the root population such as root fresh weight, root volume, root number, root length and root dry weight. The effects of SDN on the total root absorption areas and root active absorption areas of the single plants were non-significant, but it dramatically enhanced the total root absorption areas and root active absorption areas of the plant population during the tillering, heading and mature stages. In addition, SDN significantly increased the root bleeding intensity and elevated the soluble sugar and free amino acid contents of root bleeding sap. Compared to the traditional cultivation pattern (CK), SDN significantly increased root bleeding intensity at the heading stage by 4.37 and 8.90％ for early and late rice, respectively. Meanwhile, SDN profoundly enhanced the soluble sugar contents of root bleeding sap by 12.85 and 10.41％ for early and late rice, respectively. In addition, SDN also significantly enhanced free amino acid content of root bleeding sap by 43.25％ for early rice and by 37.50％ for late rice systems compared to CK. Furthermore, SDN increased the actual yield of double cropping rice mainly due to the higher effective panicle number and the larger seed-setting rate. The actual yields of early rice under SDN were higher than CK by 9.37 and 5.98％ in 2016 and 2017, and the actual yields of late rice under SDN were higher than CK by 0.20 and 1.41％ in 2016 and 2017, respectively. Correlation analysis indicated that the significant positive correlations were observed between the majority of the root indexes and the actual yield across the four different growth stages.