ABSTRACT: Freshly-isolated Symbiodinium (FIS) have been used to study cnidarian–alga symbiosis based on the assumption that their physiological performance is comparable to that of the algae in hospite. This assumption was tested with 15 species of scleractinian corals, using pulse-amplitude modulation (PAM) chlorophyll a (chl a) fluorescence to compare maximum quantum yields (F_v/F_m) of photosystem II (PSII) in Symbiodinium in hospite and after isolation in seawater. FIS from Montipora spp. exhibited rapid and dramatic decreases (by up to >95%) in PSII activity within 30 min of isolation. In contrast, PSII activities of FIS from 8 other coral species decreased by only 5 to 21% after >4 h in seawater. To investigate possible reasons for this variation, the variation in Symbiodinium genetic type (ITS-2 types), transmission modes and several physiological indices were considered. The rapid loss of PSII activity in FIS from Montipora spp. was not correlated with Symbiodinium ITS-2 type, the mode of symbiont transmission, or ionic regulation capability, nor could consistent chemical effects in host extracts be demonstrated. PSII inactivation was correlated with increased levels of reactive oxygen species (ROS), degradation of pigments in peridinin-chl a protein, and chloroplast disruption, indicating that the FIS was under physiological collapse. The specific mechanism(s) causing Montipora-associated Symbiodinium to fail during the isolation methods used here remain unknown. However, these data indicate that care should be taken when using FIS to represent Symbiodinium in hospite, especially when comparing different species of corals.