Abstract: A method and apparatus are provided for predicting failures of hard disk drive systems in set-top boxes (STBs) to enable preventative steps to be taken to prevent a disruption of services to the subscriber. One or more conditions in a STB, such as, for example, ambient temperature, available disk drive storage space, etc., are measured and a failure prediction algorithm is performed to predict the likelihood that a failure will occur. The prediction may be reported to a STB servicing entity to enable a service call to be arranged for the subscriber before services are disrupted due to a failure of the hard disk drive system.