The epidemiology of rabies in southern Africa is complex, due to a large number of vector species and the presence of at least two distinct biotypes of the virus. Our objective was to contribute to the understanding of the epidemiology of rabies in the southern African subcontinent by studying the genetic relationship of 89 rabies virus isolates from this region. In this study, we have focused on an analysis of viruses that cycle in canid host species (canid biotype) throughout South Africa and Zimbabwe. By phylogenetic analysis of the cytoplasmic domain of the glycoprotein and the non-coding G-L intergenic region, all the southern African canid viruses were found to be closely related and no apparent general distinction could be made between them. Although there was a minor degree of phylogenetic branching, with certain branches associated with cycles defined by species, location and time, the phylogenetic pattern indicated that canid rabies in southern Africa is derived from a single virus lineage, which has spread opportunistically within whatever canid host population is ecologically capable of sustaining prolonged cycles. This molecular epidemiological study presents the first comprehensive comparison of rabies viruses from South Africa and Zimbabwe and has demonstrated the need for multinational approaches towards the control of this important zoonotic disease in Africa.