Controlling costs in the acquisition of new defense systems is a major challenge in today's environment of declining budgets and rapidly changing techlogy. One of the challenges faced by program managers and cost analysts is selecting the most appropriate Estimate at Completion (EAC) method for their program. This study compares the performance of the popular index-based EAC methods with several newer nlinear, regression based EAC methods to determine whether the complex nlinear methods perform better than the simpler index-based methods. In addition, the sensitivity of the results to stage of contract completion, system type, program phase, contract type, department of Defense service component, and inflation effects are also investigated. Eighty-eight contracts were examined in this study and it was found that overall the index-based EAC methods performed significantly better than the nlinear regression based methods as measured by two criteria, the accuracy and stability of the EACs. In addition, the top performing method overall was determined to be the index-based method using the Composite Index (0.2SPIcum+0.8CPIcum). The best performing method was, however, sensitive to all of the factors investigated in the sensitivity analysis.