Abstract: Abstract：Rapid and accurate assessment of temporal and spatial variations of crop nitrogen (N) status is important to helpfarmers improve site-specific N management in sustainable agriculture. However, current studies place little emphasis on cropN estimations by taking N’s vertical distribution into consideration, leading to limited accuracy of the results. The main goalof this study was to quantitatively analyze the vertical distribution characteristics of N nutrition indices (leaf chlorophyllconcentration, leaf N content and leaf N accumulation) and SPAD value of different leaves, and determine the sensitive leafposition for N diagnosis with a portable SPAD-502 chlorophyll meter (Konica Minolta Sensing, Osaka, Japan) of summermaize. Two field experiments were conducted over two consecutive growing seasons (2018－2019) with three growth stages(large bell stage, silking stage and filling stage) at two sites (Wenxian county and Hebi city) in Henan province, North China.The same cultivar of summer maize, i. e., Zhengdan No. 958, was used during the two growing seasons. The detailed Nfertilization treatments in the two growing seasons were as follows: (i) no N application (N0); (ii) N fertilizer application rateof 75 kg/hm2, applied as urea (N75); (iii) N fertilizer application rate of 150 kg/hm2 (N150); (iv) N fertilizer application rate of225 kg/hm2 (N225); and (v) N fertilizer application rate of 300 kg/hm2 (N300). All the N nutritional resources were used asbase fertilizer prior to sowing. The SPAD value from the 1st to 12th leaf from the top on the main stem (TL1-TL12) anddifferent sites on the same leaf (20%, 40%, 60%, 80% and 100%, respectively; 0-100% are ratio of leaf length from eachmeasurement point to base to the total leaf length of each leaf) of summer maize were measured at aforementioned threegrowth stages. Meanwhile, chemical assays of these summer maize samples (N nutrition indices) were performed in thelaboratory. In total, ninety samples were used for building spectral monitoring models of N estimation. A partial least square(PLS) regression analysis was employed to quantitatively describe the relationship between the SPAD value at different leafposition and leaf sites and N indices. The prediction accuracy of the monitoring models was evaluated by comparingcoefficient of determination (R2), root mean square error (RMSE) and relative percent deviation (RPD) between the observedand predicted N nutrition indicators values. Results showed that a vertical distribution pattern of the above N parametersexisted and the pattern was bell-shaped from the upper to lower layer of the plant. The variable importance for projection(VIP) resulted from the PLS regression model were used to determine the sensitive leaves and reduce the dimensionality of theSPAD value data. The sensitive location for N nutritional diagnosis was the 60%- 80% position of the 4th fully expanded leaf(TL4) from the top of the main stem, the VIP value was above 1.40 based on the PLS model, respectively. Our resultsindicated that estimation of N status using SPAD-502 chlorophyll meter data was most effective for the 60%- 80% region onthe TL4 of summer maize leaves. The model, which considered the vertical distribution patterns of the N and the optimal leafposition and location, has demonstrated great potential to estimate the N status of the whole summer maize canopy. The resultssuggested that the sensitive leaves from the main stem of summer maize were satisfactory for inversion of the N verticaldistribution.