This paper provides a comprehensive application of Lognormal Ordinary Kriging (LOK) metamodel to optimization of a stochastic simulation problem. Kriging models were developed as an interpolation method in geology. Firstly kriging methods have been successfully used for deterministic simulation optimization problem. Recently, Kriging metamodeling has attracted a growing interest, along with other simulation optimization (SO) techniques for deterministic and stochastic systems. SO researchers have started using ordinary kriging as a robust optimization tool in stochastic SO in the last decade. In this paper SO process via lognormal ordinary kriging metamodel is proposed and a lognormal ordinary kriging algorithm is presented.