This paper deals with fault localization for a mobile hydraulic system with independent metering, but it is also applicable to other hydraulic systems. The basis for this contribution is a model-based fault detection, which generates multiple symptom patterns for various component failures using a set of parity equations (see previous publications from the authors). The symptom patterns are evaluated through different classification methods such as geometric, statistical, and artificial intelligence methods. The evaluation focuses on simple approaches that are applicable in practice. Taking into account the limitations of mobile systems, different operating points, and a variety of fault scenarios, a correct fault localization of up to 92% of detected faults is possible. After locating faults correctly, IM systems enable a range of reconfiguration modes to keep the machine's functionality and therefore rise the availability. Laboratory demonstrator tests confirm the simulation-based outcomes.