【Objective】 Soil aggregates are an essential material foundation of soil structure and an important carrier of soil nutrients. Rhizosphere is the most active region of matter and energy exchanging between crop and soil, and it differs from the bulk soil in many of physical and chemical conditions and biochemical processes, thereby influencing soil aggregation. A number of studies have been reported paying attention mostly to effects of organic manure application, tillage management and some other regulatory measures on soil aggregates during fallow seasons, rather than to differences of rhizospheric soil from bulk soil in dynamic variation of soil aggregate composition during the rice growing seasons. It is of great significance to study changes in volume of soil aggregates and contents of organic carbon and nitrogen therein in the rhizospheric and non-rhizospheric soils during the rice growing season to revelation of impacts of artificial cultivation on the changes and in-depth understanding of interactions between plant and soil and variation of soil quality.【Method】A field rhizo-bag experiment using two types of red paddy soils different in fertility was conducted in the Yingtan National Field Observation and Research Station of Farmland Ecosystem, Jiangxi Province, China. The soils were collected from the topsoil layers (0~15 cm) of two typical paddy fields near the station in March 2012. Two-layered root bags were used to separate rhizosphere from non-rhizosphere and bulk soil. During the experiment, root bags and the soils around were collected before rice transplanting and at the tillering, booting and maturing stages of rice, separately. The rhizosphere and non-rhizosphere soils were saved individually and separated into five aggregate-size fractions (>2 mm, 1 ~ 2 mm, 0.25 ~ 1 mm, 0.053 ~ 0.25 mm and <0.053 mm) using the wet sieving method. Mass fraction, percentage of aggregate destruction (PAD) and mean weight diameter (MWD) of water-stable macro-aggregates were calculated to determine stability of the aggregates. Organic carbon and total nitrogen in the aggregates were measured.【Result】 Results show that the aggregates in the low fertility soil were dominated mainly with the fraction of >0.25 mm macro-aggregates (56.2%~64.0%), and the amounts of 0.25~1 mm size was the highest. Except for the aggregates 1~2 mm in size, all the fractions of aggregates in the rhizosphere soil changed significantly in content during the rice growing season. Sampling time and interactions between sampling time and rhizospheric effect remarkably affected the amounts of the fractions of 0.25~1 mm and 0.053~0.25 mm. In the high fertility soil, <0.25 mm micro-aggregates accounted for 59.8%~72.0% of the total soil aggregates, and the 0.053~0.25 mm fraction made up the largest proportion. Sampling time affected the content of >0.25 mm macro-aggregates the most, while interactions between rhizospheric effect and sampling time did the content of >2 mm fraction the most. Compared with non-rhizospheric soils, rhizospheric soils were low in PAD, but high in MWD, which indicates that rice cultivation helps improve stability of rhizosphere soil. Contents of organic carbon (SOC) and total nitrogen (TN) were the highest in the 1~2 mm fraction of aggregates and the lowest in the 0.053~0.25 mm fraction, and obviously much higher in macro-aggregates than in micro-aggregates. The SOC content in rhizosphere aggregates during the crop maturing period did not have much difference from that before rice transplanting. But SOC contents in rhizosphere aggregates did fluctuate significantly during the tillering and booting stages of rice. The SOC content in the 0.053~0.25 mm fraction was significantly influenced by sampling time. In the low fertility soil, there was no significant difference in aggregate-associated SOC contents between rhizosphere and non-rhizosphere soils. But in the high fertility soil, SOC contents in the >2 mm fraction at the tillering stage, in the 1~2 mm fraction at the booting stage and in the 0.053~0.25 mm fraction and <0.053 mm fraction at the maturing stage were significantly lower in the rhizospheric soil than in the non-rhizospheric soil. The growth of rice decreased the content of TN in macro-aggregates. The effect of rice growth was more significant in the high fertility soil. 【Conclusion】In the low fertility soil, rhizospheric effect affects mainly the composition and stability of aggregates, while sampling time does contents of SOC and TN in aggregates. In the high fertility soil, composition of aggregates and distribution of SOC and TN vary under the joint impact of rhizospheric effect and sampling time.