Paralytic shellfish poisons (PSPs) are produced by freshwater cyanobacteria and pose a threat to human and animal drinking-water supplies. The wide range of toxin analogues (and the likelihood that further analogues remain to be discovered) means that chromatographic methods are not always reliable indicators of toxicity. Although the mouse bioassay remains the method of choice in the seafood industry, its use is increasingly being questioned on ethical grounds. The cell-based Neuro-2A neuroblastoma toxicity assay is an alternative bioassay validated for testing shellfish extracts, so it was of interest to determine its applicability with the different suite of toxin analogues produced by cyanobacteria. Cyanobacterial bloom samples from Australia, Brazil, and France were assayed using the neuroblastoma assay, liquid chromatography-tandem mass spectrometry (LC-MS/MS), high-performance liquid chromatography with postcolumn derivatization and fluorescence detection, and the Jellett Rapid Test for PSP. To assess interlaboratory variability, the neuroblastoma assay was set up in laboratories in Paris (France) and Adelaide (Australia). Neuroblastoma and chromatographic methods gave comparable results except in the case of the neurotoxic Brazilian samples: LC-MS/MS did not detect the putative new PSPs contained in these samples. Inter- and intralaboratory variability of the neuroblastoma assay was typical of biological assays but no greater than that found for interassay variability between different chromatographic determinations. The batch of Jellett Rapid Tests for PSP used did not yield quantitative results. Overall, the neuroblastoma assay was useful as a screening assay for determination of toxicity caused by saxitoxin neurotoxins in freshwater cyanobacteria, having the advantage of being sensitive to unidentified toxins that currently cannot be quantified by chromatographic means.