In clinical trials, patients should be randomly allocated to treatment and control groups that consider the balance of their prognostic factors(covariates). There are many randomization methods and stratification is popular in Korea. In stratification, patients are divided into strata based on covariates and then the patients are randomly assigned to the arms of each strata. If the number of covariates increases then the number of strata increases rapidly and the results may not be reliable when the patients are inadequate in each strata. To complement this problem Pocock and Simon (1975) suggested a new randomization method that called for minimization focusing on the balance of covariates. In this study, we compare the advantages and disadvantages, the imbalance of covariates, the power of minimization, and other randomization methods by simulation.