Zoospore production was experimentally examined during grazing on three species of algal prey (Rhodomonas, Prorocentrum minimum, or Synechococcus in single-species trials) by Neuse Estuary clonal isolates of Pfiesteria piscicida, the second known toxic Pfiesteria species (\'B\'), and the benign taxon Cryptoperidiniopsis nov. gen. (lack of bioactive compounds that cause fish stress, disease or death based on fish bioassays of multiple strains and species of this dinoflagellate from the Chesapeake Bay, the Albemarle-Pamlico, and three Florida estuaries). Prior to the experiments, strains of P. piscicida from the same clone had been maintained for 4 months in fish-killing mode (TOX - mildly toxic; fish death at 12- to 24-hour intervals with 5,000 cells/mL), temporarily nontoxic mode (NONTOX, on cryptomonad prey). An older clone of kleptochloroplastidic, \'never-toxic\' strain (NEVTOX, grown on cryptomonad prey) that had lost its ichthyotoxic activity was also compared. TOX (highly toxic; fish death at 1- to 2-hour intervals with 800 cells/mL) and NONTOX strains of Pfiesteria B were compared as well. The Cryptoperidiniopsis strain was kleptochloroplastidic and had been maintained similarly as NONTOX Pfiesteria spp. Zoospore production by all species and strains was highest with cryptomonads among the three algal prey tested. TOX strains attained lower zoospore production on algal prey than NONTOX and NEVTOX strains. Weakly toxic P. piscicida showed higher zoospore production on algal prey than the highly toxic strain of Pfiesteria B; and both NONTOX Pfiesteria spp. achieved significantly higher zoospore production on cryptomonad prey than did Cryptoperidiniopsis. The data indicate that zoospore production on algal prey differs among Pfiesteria spp. depending on the strains\' history of toxic activity, and that benign lookalike species such as Cryptoperidiniopsis should not be considered as comparable to Pfiesteria in their response to algal prey.