In this paper we study the steady-state tracking properties of sampled-data systems in the presence of structured norm-bounded time-varying uncertainty. Given a linear sampled-data system with uncertainty and a known command input, necessary and sufficient conditions for the output to track the command input to within a prescribed level epsilon greater than or equal to 0 (referred to here as E-tracking) is given. As a result, a worst-case steady-slate tracking error measure is developed. The derived measure highlights the impact of time-varying uncertainty on the steady-state tracking properties of a given sampled-data systems, and identifies those aspects of the nominal system which affect the robustness of steady-state tracking to this uncertainty.