In this paper, an advanced macromodeling tool based on z-domain orthonormal vector fitting (ZD-OVF) is proposed. This novel method provides accurate and compact models for linear subnetworks using either frequency- or time-domain tabulated data. For this purpose, a new set of orthonormal basis functions are developed. Utilizing the z-domain orthonormal basis functions improves the numerical conditioning of resulting system equations. Numerical examples are provided to demonstrate the performance and convergence properties of the new algorithm.