Diatom species distributions on remote islands are influenced by both habitat quality and location. In this study diatom distributions and water quality were explored in the Falkland Islands archipelago with reference to island floras elsewhere, using 28 freshwater sites (small ponds, shallow lowland lakes, upland tarns and several streams). Site water chemistries varied from strongly acid to circum-neutral and from low to high conductivity and water-colour. Dissolved ions co-varied with conductivity and varied strongly between sites and ion ratios indicated a strong influence of sea-salt deposition. Canonical Correspondence Analysis of the diatom and environmental data revealed three groups of sites, (1) high water-colour and low pH, (2) low conductivity and low water-colour, and (3) high conductivity and higher pH. Diatom species frequency abundances were most strongly influenced by water conductivity and pH. Diatom assemblages were composed of both cosmopolitan and regionally endemic Subantarctic species. Water chemistry was a weak predictor of diatom diversity but sites with more regionally endemic species tended also to have more cosmopolitan species. Taxonomic harmonization allowed species comparisons to be made with diatom assemblages on two islands elsewhere, Signy Island (Antarctica) and South Uist (Scottish Outer Hebrides). Detrended Correspondence Analysis of the species data from these three locations strongly discriminated the Signy Island sites and Subantarctic species characterized the Falkland Islands samples. Regionally endemic species groups within the Subantarctic zone were configured by biogeographical controls on dispersal, as well as by habitat selection and water quality. Variance partitioning showed that water chemistry and spatial (location) variables were both highly significant in accounting for similar proportions of unique variance in the combined island-species dataset. Despite overall low diversity, Subantarctic species are an important floral component and need to be incorporated into global freshwater diatom diversity patterns by using species-sensitive and carefully harmonized datasets.