The practical utility of system identification algorithms is often limited by the reliability of their implementation in finite precision arithmetic. The aim of this paper is to develop a method for the numerically reliable identification of fast sampled systems. In this paper, a data-dependent orthonormal polynomial approach is developed for systems parametrized in the δ -domain. This effectively addresses both the numerical conditioning issues encountered in frequency-domain system identification and the inherent numerical round-off problems of fast-sampled systems in the common Z-domain description. Superiority of the proposed approach is shown in an example.