This paper considers the problem of robust stability for Takagi-Sugeno fuzzy systems with time-varying delays and norm-bounded uncertainties. By choosing a more general Lyapunov-Krasovskii functional and using some slack matrix variables to express the relationship between the system matrices and the terms of Leibniz-Newton formula, a new less conservative delay-dependent condition for robust stability is derived. Neither model transformation nor bounding technique for cross terms is involved in the derivation of the stability condition. Two numerical examples are given to illustrate the effectiveness of the proposed methods.