In this paper, a wavelet-neural-based adaptive control (WNBAC) with a PI type learning algorithm is proposed. The proposed WNBAC system is composed of a wavelet neural controller and a fuzzy compensation controller. The wavelet neural control is utilized