This paper develops an optimal condition-based maintenance (CBM) strategy for a singleunit system during two-stage failure, a process that includes a normal stage and delay-time stage. Both stages are divided into the in-control state and out-of-control state. In the in-control state, the items are always produced with 100% quality, whereas in the out-of-control state, the item quality deteriorates, and minor repairs are arranged accordingly to fix this problem. This causes the optimal CBM strategy for system production to exhibit four different scenarios, where different calculations are carried out using renewal reward theory to determine the system profit. Then, the system profit is optimized by an availability-cost hybrid factor that balances the "cost per unit time" and "availability". Finally, to investigate the effects of different decision objectives on the optimization results, a sensitivity study of the cost parameter and availability-cost weight factor is conducted on the optimized results through numerical simulations. According to the simulation results, this availability-cost hybrid factor, as well as the "cost per unit time" and "availability" factors, become less sensitive when it exceeds 0.6. INDEX TERMS Condition-based maintenance, delay time, hybrid evolution factor, inspection, two-stage failure process.