We consider a so-called (R, Q) stock control system with stochastic lead time in which a quantity Q is ordered as soon as stock on hand plus on order is lower than a fixed reorder point R. In literature an abundance of models is encountered in which the parameters R and Q are optimized simultaneously. To obtain these values almost always iterative schemes are required. In this paper we elaborate the following idea: will it be possible to choose a realistic probability model for lead time demand such that explicit expressions for optimal R and Q can be obtained. As a result we obtain an explicit and rather simple expression for the optimal order quantity Q by assuming that lead time demand possesses a logistic probability density. This expression provides for small values of fixed order cost better values for Q than Wilson's lot size formula.