Abstract: Maize is an important food source globally, and the grain yield of maize is tightly associated with carbohydrate dynamics in developing ear at silking stage. However, it is unclear how carbohydrate changes is related with grain yield and its components. In order to address issues above, a 2-year field study with maize hybrid ‘Pioneer 32D79’ grown with 0, 150 and 300 kg/hm² N application was conducted at the University of Missouri Bradford Research Center, Columbia, MO, USA. Maize grain yield and its 2 major components(kernel number and kernel weight) were measured at physiological maturity when 50% of the grains exhibited black layer formation from the mid-portion of the ears. Shoot dry weight and N content were also determined at both silking and maturity. Meanwhile, soluble protein concentration was analyzed in developing ear and ear-leaf of maize, and carbohydrate concentrations including glucose, fructose, sucrose and starch were assayed in the whole ears at silking. In addition, the structural equation modeling(SEM) was used to analyze the pathways of different carbohydrate concentrations in regulating kernel number, kernel weight and grain yield across the 3 N addition gradient. Our results showed that, increasing N application rate increased maize grain yield, kernel number and kernel dry weight at physiological maturity. At silking, N applications generally increased ear dry weight, spikelet primordia, ear and ear-leaf total N and soluble protein concentrations. In developing ear, glucose and fructose concentrations of maize increased with increasing N availability, whereas sucrose and starch concentrations declined. Singular regression analysis revealed that glucose and fructose positively regulated maize grain yield and its 2 components, but starch had a negative impact on all the 3 variables. The SEM analysis indicated that kernel number and kernel weight explained 91% of the total variance in grain yield. Kernel number had greater effect on yield, with the standardized path coefficient of 0.66, while kernel weight only had a path coefficient of 0.34. Meanwhile, carbohydrate concentrations accounted for 82% and 59% variation in kernel number and 100-kernel dry weight, respectively. Monosaccharide concentration(the sum of glucose and fructose) also influenced kernel number and weight, but starch had a larger path coefficient for kernel number than kernel weight(-0.51 vs -0.31). However, no significant correlations were observed in sucrose concentration with either kernel number or weight. These results revealed that the carbohydrate composition in unpollinated ears during the critical period is a good indicator of kernel set and grain filling, as well as final grain yield. Further exploration of enzyme activities and the regulation of enzyme activities in unpollinated ears are necessary to illustrate the underlying mechanism with regard to carbohydrate dynamics in maize ear as influenced by variations in N availability. In conclusion, studies that enhance our understanding of the causal relationships of C and N dynamics in developing ears with grain yield will be critical to increase N use efficiency of maize.