The proposed identification algorithm uses output values y(KJT), sampled after J intervals of duration T, and input values u(kT), where k = 0, 1, 2,.... It is proven that the estimated output ŷ(kT) converges asymptotically to the measured output al each sampling instant. The significance of the algorithm is that it provides a basis for control based on ŷ(kT) to be performed at intervals of T rather than JT and thereby improves control performance. Simulated results show open-loop convergence of ŷ(kT) to y(kT) for J = 10.