We present a method for the prediction of the performance of a service-oriented architecture during its early stage of development. The system under scrutiny is modelled with the UML and two profiles: UML4SOA for specifying the functional behaviour, and MARTE for the non-functional performance-related characterisation. By means of a case study, we show how such a model can be interpreted as a layered queueing network. This target technique has the advantage to employ as constituent blocks entities, such as threads and processors, which arise very frequently in real deployment scenarios. Furthermore, the analytical methods for the solution of the performance model scale very well with increasing problem sizes, making it possible to efficiently evaluate the behaviour of large-scale systems.