The reliability based design optimization (RBDO) approach requires high computational cost since the uncertainties are considered.The probabilistic constraints of conventional RBDO methods can be treated in two different ways, the reliability index approach (RIA) and the performance measure approach (PMA).The optimization problem constitutes a double loop approach which employs two nested optimization loops.It consists of an outer loop for the design optimization and an inner loop which repeatedly finds the most probable point (MPP).Therefore, the computing cost of the double loop approach is extremely high.In order to overcome these difficulties, two new methods were proposed.First, sequential optimization and reliability assessment approach (SORA) was proposed to decouple a double loop process of conventional methods.Second, the single loop single vector (SLSV) method was proposed.This method can reduce the cost by elimination of the nested optimization process.The SLSV method converts the RBDO problem into a single loop deterministic optimization.However, this replacement can increase instability or inaccuracy of the convergence according to the problem characteristics.Therefore, the method may not give an accurate solution or the convergence robustness is not guaranteed.In this research, the single loop single vector method is modified by using the conjugate gradient.The conjugate gradient is calculated with the gradient directions at the most probable points (MPP) of previous cycles.Mathematical examples are solved for the verification of the proposed method.The numerical performances of the obtained results are compared to those of other RBDO methods such as the RIA, PMA, SORA and SLSV approaches.The number of function calls and gradient calls are used as measures of efficiency.The SLSV approach using the conjugate gradient is not greatly influenced by problem characteristics and improves its convergence capability.Also, the optimization cost can be reduced and a good solution satisfying the specified reliability can be obtained.