- The present study took advantage of the availability of high resolution ADS40 digital imagery to 1) systematically resample the vegetation of the Lord Howe Island Group (LHIG, excluding Ball's Pyramid); 2) conduct a numerical analysis of the floristic data; 3) map vegetation extent and the distribution of vegetation communities and 4) compare the resultant classification and mapping with those of Pickard (1983). In July 2013, a total of 86 full floristic and 105 rapid floristic sites were sampled across the island, based on a stratified random sampling design. A hierarchical agglomerative clustering strategy (Flexible UPGMA) and Bray-Curtis dissimilarity coefficient with default beta, along with nearest neighbour analysis to identify anomalous site allocations, was used to analyze the floristic data. In total 33 vegetation communities were delineated and mapped: 19 mapping units from the full floristic analysis; 7 variants identified within five of the above 19 groups; 3 mapping units from analysis of canopyonly floristic data; and 4 mapping units recognised in previous studies that are mapped but were not sampled in this survey. The resultant list of vegetation communities and non-vegetation mapping units, along with their equivalence to Pickard (1983) and DECC (2007) units, is provided. 222 plant taxa were recorded in the survey, including 47 exotic taxa. Weeds are a common component of some communities, particularly coastal strandline communities, the shrublands of the southern mountains and regenerating vegetation on landslips. The threatened plants Xylosma parvifolia, Lepidorrhachis mooreana, and Geniostoma huttonii were recorded in floristic sites. Two communities, Gnarled Mossy Cloud Forest and Lagunaria Swamp Forest are listed as Critically Endangered Ecological Communities under the NSW Biodiversity Conservation Act 2016 (BC Act). Coastal Saltmarsh, an Endangered Ecological Community listed under the BC Act, also occurs on the main island. An assessment of the conservation status of the vegetation communities recognised in the present study is provided. A number of communities warrant further consideration for listing as threatened ecological communities. Significant improvement in vegetation community attribution and spatial resolution was possible with the highresolution digital imagery, however Pickard's 1983 classification and mapping was found to be a comprehensive and accurate description of the island’s vegetation considering the imagery available at the time. This project has resulted in greatly improved accuracy of vegetation mapping linework for the LHIG. The Pickard (1983) vegetation map comprised 321 individual polygons, whereas the new map includes 1 840 polygons of sufficient accuracy to support detailed environmental planning programs, particularly within the Settlement area where spatial accuracy in the delineation of native vegetation is critical. The new vegetation communities were applied to the updated linework to complete the project. Detailed profiles were compiled of each vegetation community for which enough information was available. These profiles should prove useful in field identification of vegetation types and the assessment of their conservation status.