A new forecast-based dynamic inventory controlapproach is discussed in this paper. In this approach,forecasts and forecast uncertainties are assumed to beexogenous data known in advance at each period over afixed horizon. The control parameters are derived byusing a sequential optimization procedure. The merits ofthis approach as compared to the classical one arepresented. We focus on a single-stage and single-iteminventory system with non-stationary demand and leadtime uncertainty. A dynamic re-order point control policyis analyzed based on the new approach and its parametersare determined for a given target cycle service level. Theperformance of this policy is assessed by means ofempirical experimentation on a large demand data setfrom the pharmaceutical industry. The empirical resultsdemonstrate the benefits arising from using such a policyand allow insights to be gained on other pertinentmanagerial issues.