This research deals with an overhaul-replacement policy for a\u00a0repairable machine sold with a free replacement warranty (FRW), which will be\u00a0used for a finite horizon T and evaluated at fixed interval s. At each evaluation\u00a0point, the buyer considers three alternative decisions, i.e. (1) keep the machine,\u00a0(2) overhaul it, or (3) replace it with a new identical one. If the machine fails\u00a0during the warranty period, this is rectified at no cost to the buyer. Any failure\u00a0occurring before and after the expiry of the warranty is restored by minimal\u00a0repair. An overhaul-replacement policy was formulated using a dynamic\u00a0programming approach. The results show that overhaul rejuvenation may extend\u00a0the machine life cycle and delay the replacement decision. In contrast, the\u00a0warranty stimulates early machine replacement and in so doing increases the\u00a0replacement frequency for a certain range of replacement costs. This implies that\u00a0in order to minimize the total ownership costs over T, the buyer needs to consider\u00a0the minimal repair cost reduction due to the rejuvenation effect of an overhaul as\u00a0well as the warranty benefit due to replacement. Numerical examples are\u00a0presented for both illustrating the optimal policy and describing the behavior of\u00a0the optimal solution.<\/p>\n