The estuarine genus Pfiesteria has received considerable attention since it was first identified and proposed to be the causative agent of fish kills along the mid-Atlantic coast in 1992. The presumption has been that the mechanism of fish death is by release of one or more toxins by the dinoflagellate. In this report, we challenge the notion that Pfiesteria species produce ichthyotoxins. Specifically, we show that (i) simple centrifugation, with and without ultrasonication, is sufficient to "detoxify" water of actively fish-killing cultures of Pfiesteria shumwayae, (ii) organic extracts of lyophilized cultures are not toxic to fish, (iii) degenerate primers that amplify PKS genes from several polyketide-producing dinoflagellates failed to yield a product with P. shumwayae DNA or cDNA, and (iv) degenerate primers for NRPS genes failed to amplify any NRPS genes but (unexpectedly) yielded a band (among several) that corresponded to known or putative PKSs and fatty acid synthases. We conclude that P. shumwayae is able to kill fish by means other than releasing a toxin into bulk water. Alternative explanations of the effects attributed to Pfiesteria are suggested.