Abstract: Leaf expansion in plants is a process determined by both genetic and environmental factors. This research was aimed to quantify the variation in leaf parameters of Catalpa bungei clones under different nitrogen exponential fertilizations, establish leaf growth and leaf area models, and analyze the relationship between different nitrogen exponential fertilizations and response of foliage growth. Seedlings of four two years old C. bungei clones (No. 1-4) by tissue culture were used as experiment materials. We set four urea dose levels of exponential fertilization, i.e., control (CK), applying 6, 10, and 14g per plant, and then measured leaf length, leaf width and leaf area during the entire course of fertilization from March to August 2011 in Xiaolong Mountain Forestry Science and Technology Research Institute, Tianshui City of Gansu Province, northwestern China. Results showed that: 1) the change of leaf parameters presented an “S” curve, and leaf development was in line with Logistic growth model (P<0.01). Different nitrogen treatments and CK showed consistent impacts on leaf development parameters in June, but varied in July and August (compared with CK, leaf width increased by 31.4%-38.7% in July and 79.8%-111.2% in August , leaf area increased by 59.0%-98.8% in July and 304.4%-423.0% in August, leaf area increased by 68.2%-92.3% in August). Leaf area in N10 and N14 treatments had no significant difference in July, but the leaf area of N10 was significantly greater than N14 and N6 in August. Nitrogen of N14 was excessive in the late growth period, and the optimum amount nitrogen was 10g. 2) Logistic growth curve showed that different exponential fertilizations postponed the arrival of initial growth peak period, the maximum growth peak period and the final growth period of leaf area, and increased the maximum growth rate in July and August, but little effect of different exponential fertilizations was found on foliage expansion days. 3) Univariate, binary and ternary regressions and power functions were all able to fit the relationship between leaf length, leaf width, as well as product of leaf length × width and leaf area (R2>0.8). The best fitting indicator of leaf area model was the product of length × width, and the best fitting method was power function (for both R2>0.95). Leaf areas in the development stage for C. bungei clones under different nitrogen exponential fertilizations can be estimated by only measuring leaf length with the linear regression equation LA=-63.8014+10.8229L and power function regression equation LA=0.108L2.45 in practice (for both R2>0.93).