We construct an actuarial model of the Government Employee Pension system in Korea and evaluate the aggregate fiscal effects of the 2015 reform as well as the marginal effects of the individual measures included in the reform package. We find that our model estimates are very close to those used in the 2015 reform process by the government, which suggests that there was no evident governmental bias in estimates given the projection assumptions. Among the measures of the reform packages, the increase in the contribution rate and the cut in the accrual rate create the largest savings in the pension finance. In particular, the annual savings from the cut in the accrual rate grow in magnitude initially from a third of those from the increase in the contribution rate to a similar level eventually. Finally, results from a series of sensitivity tests and scenario analyses suggest that choosing actuarial assumptions which reflect changing population and economy in Korea can be critical in making good long-range estimates.