A simple system was used to simulate the effect of alternating toxic (paralytic shellfish poisoning toxins) and non-toxic microalgal diets on oyster feeding behaviors and rates of toxin accumulation. These experimental conditions were meant to reflect, to some extent, the incoming and outgoing fluxes of toxic algae observed at the mouth of the Penzé estuary (Northern Brittany, France). Physiological and toxicological parameters were estimated based on fluorescence measurements recorded continuously at the outlet of each experimental tank, which contained a single oyster. Q tox, this variable describes toxin uptake in oysters, it was used (instead of the toxin ingestion rate): i) in simple graphical analyses, ii) as well as in one- and two-compartment models. Results show that toxin uptake varies widely from one individual to another and is not proportional to the concentration of toxic algae in sea water. A one-compartment model with individual fluorescence recordings as “input” data gave questionable results, however, a two-compartment model was found to effectively describe contamination kinetics in oysters. Limitations of this model as well as possible improvements are discussed.