This thesis deals with the inventory problems when the values of parameters involved such as holding cost, shortage cost, ordering cost as well as demand rate are not known. Optimal inventory policies for ($t_p$,S) and (S,Q) systems are developed for the following two cases; the probability density function for each parameter is known and only the range of each parameter is known. Two different measures are defined, and for each measure the minimization of the expected values and the minimax value criteria are adapted respectively. For (S,Q) system, the results are shown to be consistent with those from EOQ. Some numerical examples are included.