Abstract: Nitrogen is the most critical input limiting rice productivity. Due to increasing fertilizer costs and environmental concerns, nitrogen use efficiency (NUE) is hotly debated in the scientific community. To explore the absorption, utilization and main traits distribution patter of nitrogen in recombinant inbred lines (RILs), a field experiment was conducted to evaluate potential NUE of rice (Oryza sativa L.). The split-plot experiment had two treatments (one without nitrogen fertilizer and the other with 150 kg.hm^-2 of nitrogen) for populations of RILs, the parents and the check line Q149. The relationship between NUE and the main agronomic characteristics of RILs population was determined using statistical correlation, clustering, principal component analyses. The results showed that the variation coefficient of NUE was larger under 150 kg.hm^-2 nitrogen treatment than under non-nitrogen treatment. Proper increase in nitrogen fertilization was beneficial to nitrogen content in rice panicle, stem and leaf, and to total dry matter weight of single plant. Significantly positive correlations were noted between dry matter production efficiency and plant high, panicle length, stem weight per plant, total dry matter weight of single plant under two nitrogen treatments. Also under both treatments, significantly negative correlations were noted between dry matter production efficiency and the contents of nitrogen in rice stem, leaf and panicle. There were positive correlations between grain production efficiency and grain weight per plant, seed setting rate, thousand-seed weight, total number of grains per panicle, and spike length. Equally, there were negative correlations between grain production efficiency and stem weight per plant, leaf nitrogen content, leaf weight per plant and total amount of nitrogen per plant. Stepwise regression analysis indicated that nitrogen content in stem and panicle, and stem weight per plant had significant effects on dry matter NUE. However, the effects of number of panicles per plant, number of grains per panicle and seed setting rate on grain NUE were more significant in the two treatments. Principal component analysis showed that the nitrogen content in rice plant, especially in stem, decreased with increasing of NUE. Therefore, in low-nitrogen field, it was necessary to select heavy-panicle type variety that was high in stem weight (including long panicle length, more grains per panicle and high seed-setting rate) in breeding programs. Furthermore, lower nitrogen content in stem and panicle (especially in stem) benefited the breeding for high NUE. Based on the study, 7 rice lines with high NUE (e.g., Q149) and 6 lines with low NUE (e.g., Q114) were selected for special germplasm in rice breeding projects.