Available bioassays for detecting brevetoxins in marine seafood were compared, and correlated against analytical HPLC procedures and the AOAC i.p. mouse bioassay.   Radioligand synaptosome binding assays, radioimmunoassays, enzyme-linked immunoassays, Gambusia affinis fish bioassays, and electrophysiology were utilized as potential substitutes for mouse bioassay.  Assays were performed on both natural field-collected specimens before, during, and after a red tide, from laboratory-induced toxic specimens, and from homogenates of seafood sources artificially "spiked" with brevetoxins.  P. brevis cells or toxin in seawater could be detected using any of the assays utilized.  ELISA and RIA were most sensitive in identifying low amounts of brevetoxin.  Several new formats were explored, specifically designed for water use, for use as a "presumptive" assay, or for the rigors of routine tissue testing Synaptosome binding assays and electrophysiology provided the most sensitive methods of identifying very potent toxins.   Gambusia affinis bioassay proved most useful for rapid screening of toxic fractions, while HPLC methods proved relatively insensitive to toxins with little carbon-carbon unsaturation, and gave no indication of potential toxicity.  Biological matrices of shellfish or fish tissue proved problematic for ELISA and HPLC, while RIA and mouse bioassay could detect toxin in these matrices. Two unexpected results were the identification of a new hemibrevetoxin (uncharacterized structurally, Btx-H) from Mote Marine Laboratory cultures, and the discover of irreversible brevetoxin binding to fish tissues.  Spectral results are presented for the new hemibrevetoxin, and data is presented for the latter irreversible binding. The ultimate disposition of brevetoxin in fish tissues is not known.  That is to say, it is not known if toxin is metabolized to non-detectable forms or if it is sequestered in tissues to non-extractable forms.