We examined cell production in P. piscicida given fish versus algal prey. Clonal culture with demonstrated lethality to fish prey was isolated from the 1998 toxic Pfiesteria outbreak on the Neuse Estuary, North Carolina, USA., and was maintained in toxic (TOX, with live fish prey) and [temporarily] nontoxic modes (NONTOX, with cryptomonad prey) for 4 months prior to the experiments. A second clonal Neuse culture, isolated in 1997, was no longer toxic to fish and was grown on cryptomonads as a \'never-toxic,\' kleptochloroplastidic strain (NEVTOX). Zoospores encysted when prey were allowed to deplete. Cysts were maintained for 1-26 weeks, with excystment triggered at initiation of experiments by adding fish or algal prey. The history of access to live fish strongly influenced subsequent cell production. When given fish prey, TOX zoospores had significantly higher rates of excystment and cell production than NONTOX and NEVTOX strains that previously had been grown with algae. When given algal prey (f/100 media), cell production by the NEVTOX strain generally was signficantly higher than by NONTOX (but potentially toxic) P. piscicida, with lowest cell production by the TOX strain that previously had been maintained in toxic fish-killing mode. Nontoxic zoospores became never-toxic - i.e., no longer showed attraction to fish or fish-killing ability - after 2-4 months on algal prey without access to fish. Under N or P enrichment with algal prey, highest and most rapid cell production was shown by the N-enriched NEVTOX strain, and a non-prey-mediated response was shown by the P-enriched TOX strain. The data from these experiments and similar research with other clones indicate that in culture, P. piscicida generally loses its fish-killing ability if maintained for more than short periods on algal prey; and that meaningful insights about the ecology of toxic P. piscicida cannot be obtained from research on never-toxic strains.