In this talk, we address the exponential stability analysis for neutral stochastic systems with multiple time-varying delays. By constructing a novel Lyapunov-Krasovskii functional, we derive a new delay-dependent stability criterion formulated in terms of linear matrix inequalities (LMIs). A key contribution of our work is that the resulting conditions involve fewer matrix variables, which significantly reduces the computational burden. I will highlight that our approach is particularly effective in reducing conservatism for systems with time-invariant delays. Finally, the effectiveness and advantages of our proposed method will be demonstrated through three numerical examples.