Photosynthetic performance and cellular chlorophyll a (chl a) content were studied in the mixotrophic dinoflagellate Fragilidium subglobosum during transition from phototrophic to mixotrophic growth and back again. Experiments were conducted at an irradiance of 50 μE m^-2 s^-1 and, for mixotrophic growth, in the presence of Ceratium tripos or C. lineatum. After 4 to 8 d in mixotrophic culture, cells of F. subglobosum had reduced both their cellular chl a content to 47% and the photosynthetic performance to ~42% of that found in phototrophic cultures. Net production of chl a in phototrophic and mixotrophic F. subglobosum cells was the same, indicating that feeding in F. subglobosum does not repress chl a production in mixotrophic cells. Thus, the reduction in cellular chl a content observed in mixotrophic F. subglobosum cells was caused by an increase in growth rate due to feeding. In cultures close to steady state, the photosynthetic performance at low prey concentrations was similar to values obtained by monocultures of F. subglobosum. However, above a certain prey concentration, photosynthetic performance decreased exponentially with an increase in prey concentration. The application of the traditional ¹⁴C labeled HCO₃ ^- technique to measure photosynthesis in mixotrophically grown cells underestimated rates by 5 to 12%, probably due to fixation of CO₂ produced from respiration of ingested carbon. In phototrophic and mixotrophic F. subglobosum cells 46 and 28% of the assimilated carbon is used for respiration, respectively. This indicates that F. subglobosum needs more energy to synthesize, maintain and run the photosynthetic apparatus than the heterotrophic apparatus involved in prey capture, digestion and assimilation.