Abstract-In order to explore the responses of soil enzyme activities and microbial community compositions to long-term nitrogen (N) addition in both bulk soil and microaggregate of chestnut soil, we conducted a 7-year urea addition experiment with N treatments at 6 levels (0,56, 112, 224, 392 and 560 kg N ha -1 yr -1 ) in a temperate steppe of Inner Mongolia in China. Soil properties and the activities of four enzymes involved in carbon (C), nitrogen (N) and phosphorus (P) cycling were measured in both bulk soil and microaggregate, and phospholipid fatty acids (PLFAs) were measured in bulk soil. The results indicated that: 1) in bulk soil, N addition significantly decreased β-1,4-glucosidase (BG) and leucine aminopeptidase (LAP) activities at the treatment amounts of 224, 392 and 560 kg N ha -1 yr -1 , and obviously suppressed β-1,4-N-acetylglucosaminidase (NAG) activity at the treatment amount of 560 kg N ha -1 yr -1. N addition enhanced total PLFAs (totPLFAs) and bacterial PLFAs (bacPLFAs) at the treatment amounts of 392 and 560 kg N ha -1 yr -1 , respectively, but fungal PLFAs showed no response to N addition. The activities of BG, NAG and LAP were positively correlated with soil pH, but negatively correlated with the concentration of -N; 2) in microaggregate (53-250 μm), the activities of BG, NAG and AP showed no response to increased addition of N, but the significantly decreased LAP activity was observed at the treatment amount of 392 kg N ha -1 yr -1 . These results suggested that enzyme activities were more sensitive to N addition than PLFA biomarkers in soil, and LAP activity in microaggregate may be a good indicator for evaluating N cycle response to long-term N addition.