In this paper, the extended application of an approximating optimization formula, which is first proposed to simplify the optimization of networks with a single traffic class in the literature, in QoS-supporting networks is analytically justified and numerically investigated. Based on the formula, an adaptive optimization scheme, named QATC (QoS-supporting adaptive transmission control), is proposed for IEEE 802.11 to optimize the system performance, while providing service differentiation among traffic classes. The proposed scheme works around the difficult station number estimation. It utilizes information acquired from channel sensing to adjust the packet transmitting, adaptively towards achieving an invariable system object for different network conditions. Furthermore, the scheme to optimize the multi-rate networks is also exploited. Simulation results show that QATC has an effective ability to optimize the operation of networks adaptively in various conditions, and it can greatly improve the system throughput compared with the standard IEEE 802.11 MAC and related enhancements. Moreover, the system throughput, achieved by the scheme, is very close to the theoretical optimal value.