The production and processing of cotton is a major source of foreign exchange for the economy of Pakistan. The majority of cotton fiber produced in the region comes from the tetraploid Gossypium hirsutum, although some is still produced from the cotton species native to the region, the diploid G. arboreum. Since the early 1990s, cotton production in Pakistan and northwestern India has been adversely affected by cotton leaf curl disease (CLCuD). The disease is caused by single-stranded DNA viruses of the genus Begomovirus (family Geminiviridae) in association with a specific satellite, Cotton leaf curl Multan betasatellite (CLCuMuB). At this time only a single virus, Cotton leaf curl Burewala virus (CLCuBuV), is associated with CLCuD across most of Pakistan. This virus is resistance breaking, overcoming resistance to the previous begomoviruses/satellite complex that was introduced into cotton by conventional breeding. At this time there are no commercially available G. hirsutum lines that are resistant to CLCuBuV/CLCuMuB. However, all lines of G. arboreum are “immune” to CLCuD and plant breeders have long been trying to introduce the “resistance” from this species into the more desirable G. hirsutum lines. In addition, recently two lines of G. hirsutum originating from France (cvs. Dominique and Haiti) have shown promise in field screening for resistance against CLCuD. The study described here was designed to investigate the nature of the resistance of G. arboreum cv. Ravi and the French G. hirsutum cultivars, Dominique and Haiti, using whitefly-mediated and graft inoculation of the CLCuD virus complex. Additionally the possibility of using biolistic inoculation of viral DNA was investigated as a possible means of experimentally introducing the virus complex causing CLCuD into cotton. In large scale field screening of G. arboreum cv. Ravi over a period of two years, no symptoms of virus infection were detected under inoculation pressure conditions where 79-89% of the susceptible control (G. hirsutum cv. CIM 496) plants were symptomatic. Rolling circle amplification/polymerase chain reaction xiii (RCA/PCR) diagnostics, the most sensitive diagnostic method available to detect geminiviruses in plants, did not detect either CLCuBuV or CLCuMuB in field grown G. arboreum cv. Ravi plants; consistent with the idea that G. arboreum is immune to the virus complex. However, graft inoculation with scions from CLCuD affected G. hirsutum showed firstly that the virus complex can move systemically in the plant and that G. arboreum can respond to virus infections by the production of symptoms. Surprisingly, in a few cases, the disappearance of established symptoms was seen following removal of the graft. In all graft inoculated Ravi plants, after removal of the graft, newly emerging tissues were non-symptomatic and no virus could be detected. These results show that, rather than being immune, G. arboreum is highly resistant to the CLCuD complex and has a high virus/satellite threshold for the induction of symptoms, which whitefly inoculation likely is not able to achieve. The low virus levels detected in G. arboreum suggest that possibly the resistance targets virus/satellite replication and, without a continual source (such as from a graft), the virus/satellite complex is rapidly lost. In small-scale, glasshouse-based insect transmission studies, plants of G. hirsutum cvs. Dominique and Haiti remained symptomless under conditions where all G. hirsutum cv. CIM 496 plants became infected. Graft inoculation showed the Dominique and Haiti plants to be susceptible but showing only mild symptoms, slightly higher than in grafted G. arboreum cv. Ravi plants. The virus/satellite levels in such plants were lower than in the susceptible control but higher than detected in G. arboreum cv. Ravi. Upon removal of the graft, newly developing leaves did not show symptoms and no virus/satellite could be detected. The response to infection seen in G. hirsutum cvs. Dominique and Haiti very much mirrors what was seen for G. arboreum cv. Ravi. Recovery from infection has, for other viruses, been shown to be an RNA interference phenomenon and the results are discussed in light of this possibility. G. hirsutum cvs. Dominique, Haiti, Coker and S-12, as well as G. arboreum cv. Ravi plants were biolistically inoculated with cloned CLCuBuV/CLCuMuB, Cotton leaf curl Kokhran virus (CLCuKoV; a begomovirus prevalent in cotton in Pakistan in the 1990s)/CLCuMuB and with RCA products from field-infected G. hirsutum cv. CIM 496 plants shown to be infected with CLCuBuV/CLCuMuB. Only a small number of Coker and S-12 plants, inoculated with cloned xiv CLCuKoV/CLCuMuB, became infected and showed the typical symptoms of CLCuD. Overall the findings indicate that G. hirsutum cvs. Dominique and Haiti harbor a useful resistance to the virus(es) causing CLCuD which should be used for introgression into elite cotton varieties. The results obtained with G. arboreum cv. Ravi indicate that, rather than being a non-host, this harbors extreme resistance to the viruses causing CLCuD and further efforts should be made to characterize the molecular basis for the resistance. Finally the biolistic studies indicate that this can potentially be a useful method for experimentally introducing begomoviruses/satellites into which should be investigated further.