A reliability prediction study has been carried out using failure data from the gas turbine system at a combined cycle power plant in Indonesia. From this study, the prediction value of the equipment reliability of the gas turbine subsystem was obtained along with the value of Failure Cost (FC), Component Risk (CR), and Maintenance Cost (MC) which is then processed further to produce a Benefit to Cost Ratio (BCR) which is useful for preparing priority rankings for the allocation of very limited maintenance budgets. This research also leads to a conclusion that the more the failures experienced by the equipment in one observation period, the lower the predictive value of its reliability. In addition, it is also discovered the fact that subsystems with the same number of failures but with different times of failure will produce different reliability values. And lastly, the subsystem risk rating obtained from Component Risk (CR) does not necessarily become a budget priority rating; this rating can change when the risk rating is combined with Maintenance Cost (MC) to produce a more objective budget priority rating with broader considerations. The methodology that combines the prediction of equipment reliability with budget prioritization in this research can also be used in other power generation plants that predominantly manage physical assets in order to obtain optimal asset life cycle management without compromising the performance of the equipment in the power plant due to limited maintenance budgets.