Growth in global food demand and a decline in wild fish catch has seen aquaculture become the world's fastest growing food production sector globally. In Australia and internationally, marine harmful algal blooms (HABs) caused by certain dinoflagellates and diatoms have increased in parallel with aquaculture growth and ocean climate change, and have led to the uptake of biotoxins in seafood and deaths of marine life. Molecular genetic tools allow insight into the genetic basis of marine biotoxins and HAB development, and provide the potential to design new detection technologies, some of which can be rapidly conducted on site. We have isolated and cultured strains of most marine HAB species from Australian marine waters, and investigated their toxin production via high resolution LC-MS at a purpose-built facility established at the Sydney Institute of Marine Science. These include species of the dinoflagellates Alexandrium (Paralytic Shellfish Toxins), Prorocentrum minimum, a common HAB species associated with eutrophication, Prorocentrum species, producers of Diarrhetic Shellfish Toxins, Gambierdiscus spp, associated with Ciguatera Poisoning, Ostreopsis spp, producer of Palytoxins and the diatom Pseudo-nitzschia cf. cuspidata, the most common source of Amnesic Shellfish Toxins in Australia. We have collected long term, weekly eDNA datasets from 13 oyster aquaculture estuaries and a mussel farm, and used the data to model environmental factors associated with some HAB species and biotoxins, using rapid molecular assays we have developed for key species and key genes (ie sxtA, Dinophysis, Pseudo-nitzschia). We compare results of qPCR quantification assays to information collected from other sources including light microscopy-based detection, toxin detection using LC-MS and molecular barcoding based on the detection of the V4 region of the 18S rRNA gene. Genetic methods hold considerable promise due to their potential for automation and on site use, in turn reducing the regulatory burden requiring seafood analysis via LC-MS.