In this paper we compare measured rates of C-fixation obtained using in situ deployment of a fast repetition rate fluorometer (FRRF) applying two different calculation protocols and a waterPAM fluorometer. The measurements were carried out in the large, shallow Lake IJsselmeer (The Netherlands). Turbidity appears to be the most important factor in explaining photosynthetic parameters. Non-photochemical quenching (NPQ) in the antenna was observed, and the degree of NPQ was positively correlated with the irradiance and mixing conditions. However, NPQ did not affect the functional cross-section. Conversion of photosynthetic electron transport rates (ETR) into rates of carbon fixation requires knowledge of two or three parameters (depending on the protocol and instrument used) to calculate C-fixation from ETR: the electron yield (È e , i.e. the reciprocal of the quantum requirement of PSII), the photosynthetic quotient (PQ) and the number of PSII (n PSII ). Using normally assumed values for these three parameters resulted in overestimation of the measured rate of C-fixation but, in all cases, ETR was linearly related to it. The degree of overestimation was rather constant, despite considerable changes in phytoplankton composition. Spectral correction of fluorescence data caused a reduction of 30% or an increase of 44% in the estimated C-fixation, depending how C-fixation was estimated. We describe a method that allows estimation of the È e /PQ ratio based on a comparison of the quantum efficiencies for C-fixation and PSII and show that, with a single and realistic value for this ratio, primary production could be accurately predicted at different times of the year.