The potyvirus Papaya ringspot virus (PRSV) is widespread throughout the world in cucurbits (such as zucchini, watermelon, pumpkin etc) and papaya (papaw). There are two serologically indistinguishable strains of PRSV, which can only be differentiated on the basis of host range. PRSV-P is able to infect both papaya and cucurbits whereas PRSV-W only infects cucurbits. Both infections drastically reduce the yield and market quality of the fruit. Australian isolates of PRSV-P and -W are very closely related and there is evidence that PRSV-P arose by mutation from PRSV-W. The aim of this project was to investigate the molecular basis of the host range difference between Australian isolates of PRSV-P and -W. The close relationship between Australian PRSV-P and -W isolates at the molecular level made this an ideal system to investigate molecular host range determinants through the development of full-length infectious cDNA clones. Initially, the complete genomes of PRSV-P and -W were each incorporated into two overlapping clones; one included the CaMV 35S promoter fused to the 5' one third of the PRSV genome and the second included the 3' two thirds of the genome (including a 33 nucleotide poly(A) tail) fused to a CaMV35S terminator. Full-length clones could not be obtained from subcloning of these fragments due to apparent toxicity in E.coli. Several approaches were subsequently undertaken to overcome this problem. In an attempt to prevent transcription of potentially toxic sequences, a plant intron (St-Ls1 IV2 intron) was engineered into the first coding region (P1) of the PRSV-W genome. Although clones were obtained using this strategy these could not be effectively maintained in E.coli. An alternative strategy involved subcloning of the genome into a low copy number vector, pACYC177, to minimise expression of toxic sequences. Again this resulted in clones that produced very small colonies, which were hard to culture and which gave very low plasmid yields. These plasmids were also difficult to maintain in E. coli. A final, successful strategy was developed using overlapping long distance PCR (OE-LD PCR) to generate full-length infectious PCR products of both PRSV-P (rPRSV-P) and -W (rPRSV-W) incorporating a CaMV 35S promoter and terminator. Infectious PCR products of both strains were inoculated onto squash cotyledons in vitro by microprojectile bombardment and subsequently mechanically inoculated to squash with greater than 86% efficiency. RPRSV-P subsequently infected papaya with 96% efficiency while, as expected, rPRSV-W was unable to infect papaya. Once a system for generating infectious clones was developed, both sequence analysis and recombination of infectious clones was utilised to investigate the underlying host range mechanism. The complete genomes of PRSV-P and -W were sequenced and compared to each other and to five full- length sequences of overseas PRSV isolates that were available. Sequence analysis confirmed the close relationship between the Australian PRSV isolates (97.8% nucleotide and 98.4% amino acid identity over the whole genome), supporting the mutation theory between both Australian and Asian P and W pairs. However, there was no consistent amino acid difference over the whole genome that correlated with host range or a single site that could be implicated, suggesting that the mutation and possibly the position of the mutation is different at least between Asian and Australian isolates and potentially differs at each mutation event. To better localise the P/W mutation within the PRSV genome, five different recombinant hybrid PRSVs (rhPRSV1-5) were generated in which 5', middle or 3' regions of the PRSV-P and -W genomes were exchanged. Infectivity of all hybrids was confirmed in squash, however, only hybrids including the 3' third of the PRSV-P genome were able to infect papaya, suggesting that this region encodes the papaya host range determinant. The region implicated encodes the genome-linked protein (VPg), NIa protease, replicase (NIb), coat protein (CP) and 3' UTR. While further identification of the host range determinants was not possible due to time constraints, based on studies with other potyviruses, there is a strong basis for implication of the VPg. Sequence analysis identified only 2 amino acid differences between the VPg of Australian PRSV-P and -W isolates in regions previously implicated in pathogenicity. These will be targeted for mutagenesis in ongoing studies. Identification of the genes/sequences involved in the determination of host range in PRSV will provide valuable information as to the sequence of events that lead to infection and will lead to a better understanding of the significance of changing hosts in the molecular evolution of PRSV, an essential requirement for the development of long-term sustainable control strategies against PRSV.