Farming is faced with the problems of excessive fertilizer use and the reduction of nutrient use efficiency. Phosphorus is one of the macronutrients necessary for plant growth and development. However, phosphorus is usually immobilized by organic substances or oxidized surfaces of iron-aluminum, and exists in the form of organic or inorganic phosphorus, which makes uptake from most soils difficult. It is well known that acid phosphatase increases the use efficiency of organic phosphorus by releasing inorganic phosphorus from organic phosphorus such as phytate. AtPAP15 encodes a purple acid phosphatase, which has been observed to have phytase activity, in Arabidopsis thaliana. AtPAP15 was over-expressed in transgenic soybean AP15-1 and its expression product was secreted into the rhizosphere because of the presence of a carrot extension gene signal peptide, and thus enhanced use of soil organic phosphorus. Previously, there have not been many reports of research concerning the effects of such nutrient-efficient transgenic plants on soil ecology. In this study, the effects of P-efficient transgenic soybean AP15-1 on soil ecology, especially on the rhizosphere microbial community, were investigated. Transgenic soybean AP15-1 and its receptor YC03-3 were planted in vent-netting greenhouse in 3 m × 5 m experimental plots with three replicates. Soybeans were planted in two successive spring and autumn seasons, and no manure, herbicide or pesticide was used. To ascertain if phytase activity in root exudates of transgenic soybean AP15-1 affects the soil microbe ecosystem, the rhizosphere and non-rhizosphere soil of AP15-1 and its receptor YC03-3 were collected at seedling, florescence and mature developmental stages and analyzed by plate culture count and BIOLOG-EcoPlate. The number of colony forming units (CFU) indicated that the number of culturable bacteria, fungi and actinomyces varied with seasons and development stages. There were no apparent differences in numbers of bacteria, fungi and actinomyces among rhizosphere and non-rhizosphere soil of AP15-1 and YC03-3. The absorbance of BIOLOG-EcoPlate at 590 nm and 750 nm was recorded using microplate spectrophotometer and the average well color development (AWCD) values were calculated. The curves of AWCD versus incubation time showed that the carbon metabolic activity of microbes in rhizosphere soil was higher than that in non-rhizosphere soil from the same season. However, statistical analysis indicated that no significant differences in carbon metabolic activity and carbon source preference were observed among rhizosphere and non-rhizosphere soils of AP15-1 and YC03-3. Statistical analysis of the calculated microbial abundance and diversity index also exhibited no distinct differences among rhizosphere and non-rhizosphere soils of AP15-1 and YC03-3. Principal component and clustering analyses revealed that the difference between the rhizosphere of AP15-1 and YC03-3 was less than that between rhizosphere and non-rhizosphere soil from the same season and at the same plant developmental stage. The effects of season and developmental stage on the community diversity of rhizosphere microbes were greater than that resulting from the genotype differences between AP15-1 and YC03-3. In conclusion, these results demonstrated that the P-efficient transgenic soybean AP15-1 planted in spring and autumn had no significant impact on the soil microbial numbers and community structures.