The seminar work of Gupta and Kumar [1] showed that multi-hop wireless networks with capacity scalable with the number of nodes, n, are achievable in theory. The transport capacity scales as \Theta \sqrt n , while the capacity scales as \Theta (n) . A subsequent study [2], on the other hand, showed that the capacity of IEEE 802.11 networks does not scale with n due to its carrier-sensing mechanism. This prior work, however, has not considered the use of power control. The main contributions of this paper are three-folds: 1) we provide an analytical framework for deriving the design requirements of adaptive power control strategies; 2) we demonstrate that 802.11 networks are scalable with power control; 3) however, an enhanced MAC protocol called Selective Disregard of NAVs (SDN) can achieve substantially higher capacity with an adaptive power control scheme; in particular, adaptive power control allows SDN to achieve capacity within 75% of the theoretical optimal capacity of infrastructure-mode wireless networks. A reason why adaptive power control works well is that it takes into consideration the fundamental mutual-interference relationships between links in the vicinity of each other, and adjust their relative transmit powers to reduce these interferences to a large extent that is possible theoretically.