In one embodiment, a method for parallel routing of a circuit design is provided. Placement of a netlist of the circuit design is determined for a target device. A plurality of regions of the target device is defined. Each region of the plurality of regions is assigned to a respective set of processors, each set including at least one processor. Global routing of nets of the netlist on the target device is performed. The global routing of each net restricts the net to one or more possible routes through a corresponding subset of the plurality of regions. Local routing of the netlist is concurrently performed within the plurality of regions using the respective sets of processors. Within each region, the local routing of the netlist is performed exclusively by the respective set of one or more processors.