In this talk, I will present and discuss a cluster-based parallel version of our hybrid feature selection algorithm FortalFS. The design of this parallel solution was based on the Master-Slave design pattern and requires only 2(n-1) messages (remote method calls) to be exchanged during the whole execution with n processors (clusters). A Java RMI (Remote Method Invocation) implementation of this parallel algorithm was developed and evaluated on a local network over four UCI datasets. For all datasets, we were able to achieve near linear speedup.