The focus of this study is on investigating worker assignment policies in dual resource constrained (DRC) job shops which include features such as the presence of heterogeneous resources and worker transfer delays between different departments. Previous studies on DRC systems have not modelled a shop which includes these design features simultaneously, although both of them would be present in many actual DRC environments. Several worker assignment policies that can be expected to work well in such environments are proposed and tested. Results indicate that in most DRC shops that have heterogeneous resources and finite transfer delays, decentralized assignments which delay the transfer decisions improve system performance. Additionally, two different task selection rules are shown to be very robust in giving good performance across a wide range of DRC environments modelled in the study.