This paper proposes an adaptive virtual machine (VM) assignment scheme for multi-tenant data center networks. In multi-tenant data centers, tenants submit their resource requirements and the data centers provide VMs which are assigned to physical servers. These VMs communicate with each other to execute distributed processing. The amount of traffic communicated by VMs is very large and this is one of big issues in data center networks. In order to reduce the amount of traffic injected into data center networks, we need an appropriate VM assignment strategy which considers traffic between VMs. Furthermore, VMs are dynamically assigned to and removed from physical servers according to resource requirements of tenants, and thus we should consider such dynamic behavior. The proposed scheme provides an VM assignment strategy which reduces the amount of traffic injected into data center networks, taking into account dynamic behavior of the data center networks. Through simulation experiments, we demonstrate that the proposed scheme reduces the network load efficiently.