The article studies the admissibility stability issue of Takagi-Sugeno fuzzy singular(TFS) systems with time-varying delay. First, negative-definiteness constraints for nth-order matrix-valued polynomials are established based on Finsler's Lemma. A new reciprocally convex inequality(RCI) is developed according to the proposed negative-definiteness method. Then, an improved state decomposition Lyapunov-Krasovskii(L-K) functional is constructed based on the decomposed state method. The improved L-K functional is augmented by considering more information on TFS system states and the delay-dependent matrix. Based on the novel RCI and the L-K functional, a less conservative admissibility criterion is derived. Finally, a well-studied numerical example is performed to reveal the superiority and validity of the proposed admissibility criteria.