PARUS is a parallel programing framework that allows building parallel programs in data flow graph notation. The data flow graph is created by developer either manually or automatically with the help of a script. The graph is then converted to C++/MPI source code and linked with the PARUS runtime system. The next step is the parallel program execution on a cluster or multiprocessor system. PARUS also implements some approaches for load balancing on heterogeneous multiprocessor system. There is a set of MPI tests that allow developer to estimate the information about communications in a multiprocessor or cluster.