Abstract: Precise estimation of maize yield potential would enhance our understanding of crop development and yield formation and thus improve yields. Many crop models have been used in estimating yield potentials, among which the FAO-AEZ model has been shown to be efficient and accurate. However, some parameters of the FAO-AEZ model such as harvest index are set constant which limit estimation precision. Thus in this paper, we developed an optimized FAO-AEZ model to increase yield estimation accuracy. Field experiments were conducted at three ecological zones (Qian'an County, Gongzhuling City and Huadian City) in Jilin Province in 2005-2013. Dry matter accumulation at several developmental stages and grain yield of maize were measured, and then an optimized calculation method integrated with traditional FAO-AEZ model to evaluate maize yield potential. With this new method, harvest index had become a dynamic factor that was adjustable in line with maize development. Moreover, an analytical method for dry matter accumulation was developed under super-high yield production of maize and the model parameters calculated using Logistic model. The results showed a high correlation (P < 0.01) between maize yield and dry matter accumulation parameters. Then using data from independent field experiments in 2012 and 2013 to test the optimized model, low error (both absolute error and relative mean error) was noted. Using nonlinear optimization theory, the optimized model was applied in analysis of potential maize yield in Qian'an County and Huadian City. Compared with traditional FAO-AEZ model, the calculated yield potentials by the optimized model were higher by 17.5% in Qian'an County and by 16.1% in Huadian City. Furthermore, we estimated minimum planting density and dry matter accumulation at different developmental stages using a target yield of 15 000 kg·hm^-2. The minimum planting density was 7.7 10⁴ plants·hm^-2, 8.2 10⁴ plants·hm^-2 and 7.9 10⁴ plants·hm^-2 in Qian'an County, Gongzhuling City and Huadian City, respectively. The optimized FAO-AEZ model provided the scientific basis for decision-making before planting and for growth period regulation in order to have super-high-yield maize production. The established model and the results of the analysis could be used to estimate yield potential, high-yield and super-high-yield cultivation in Jilin Province and other regions for maize and other crops.