Machines operating in sandy environments are damaged by the abrasive action of sand particles that enter the machine and become entrapped between components and contacting surfaces. In the case of the military services the combination of a sandy environment and the wide range of tasks to be fulfilled results in extreme and uncertain operating conditions. All of this hinders the ability to establish efficient maintenance strategies prior to deployment and increases the risk of mechanical failure. To prevent such problems, it would be desirable to perform maintenance based on the prevailing condition of both the components and the environment. By monitoring the loading situation as well as the characteristics of the sand particles, the wear of components is quantified, allowing maintenance to be performed when necessary. The development and implementation of such a predictive maintenance concept requires knowledge of the operational and environmental conditions and how they relate to the principal wear mechanisms. Based on previously established relationships between the abrasive particles and the resulting abrasive wear, the current work focuses on the implementation of these results into a predictive maintenance concept for vehicles that operate in a sandy environment. For this, the local parameters that govern the wear mechanism, such as the normal forces and sliding distances need to be linked to machine usage parameters including the type of terrain and the driving distance. The proposed concept is demonstrated using a case study on the sprockets of a military vehicle, where the sprockets wear progressively during use of the vehicle due to the abrasive action of sand. The predictive maintenance concept is shown to support the determination of maintenance intervals under a range of usage profiles and sand varieties.