In this paper, Adaptive Noise Cancellation using Dynamic Fuzzy Neural Networks (D-FNN) algorithm is attempted. The D-FNN algorithm is a hierarchical on-line self-learning algorithm based on Extended Radial Basis Function Neural Networks. The salient features of the algorithm are: (1) It provides an efficient learning method; (2) The weights are modified by the Recursive Least Square method; (3) No iteration is needed. Simulation studies and comparisons with other algorithms show that the D-FNN algorithm is superior in terms of simplicity of structure, learning efficiency and performance in canceling noise adaptively.