SUMMARY: Methods for the laboratory maintenance and purification of cultures of Prymnesium parvum are detailed. The upper limit of temperature for the growth of the organism is 30". Light exerts an augmenting effect on the production of the toxin by Pr. paruum. Rapid and economical procedures for the bioassay of the toxin are based on the use of Gambusia minnows or tadpoles. Cell-free centrifugates of the cultures contain heat-labile toxic material which is non-diffusible through cellophan, sensitive to oxidizing agents, and reversibly inactivated by mild acidity. The toxin is rapidly inactivated by ubiquitous bacterial species (Bacillus subtilis and Proteus vulgaris). Charcoals, clays and calcium sulphate are efficient adsorbents of the toxic material. The concentration levels of toxin in cultures of Prymnesium parvum appear to reflect a dynamic equilibrium between toxin destruction and production ; a similar equilibrium may prevail in nature. ,4 phytoflagellate toxin was first implicated in an occurrence of mass mortality among fish in Holland by Liebert & Deerns (1920). Subsequent study of a similar occurrence in Denmark by Otterstroem & Steeman-Nielsen (1939) identified the responsible organism as Prymnesium parvum Carter, and con- firmed that the toxic effect was caused by an extracellular thermolabile toxin. Reich & Aschner (1947) described the widespread occurrence of this organism in Israel and proposed measures for its control in fishponds. The present communication describes methods for the maintenance and purification of laboratory cultures of Prymnesium spp. and procedures for the bioassay of the toxin. An analysis of the effects of certain environmental factors on the production and persistence of the toxin in pond water has been attempted. The experiments show the extracellular occurrence of the prym- nesium toxin and suggest that it may be a protein. EXPERIMENTAL Bioassay of the toxi% of Prymnesium parvum Estimations of toxin content of culture fluids and of pond water were made by using Gambusia minnows or tadpoles of Rana or Bufo as the test organism. These species were readily available and could be handled in relatively small volumes of fluid. The tests are carried out as follows. For tadpole tests, five animals are immersed in 5 ml. volumes in 15 mm. diam. test-tubes. Cell-free centrifugates of the toxic samples are used for the assay. The dilution medium was sodium chloride solution (0.16 yo), 9 volumes + Sorensen's phosphate