AbstractMany recent models for retrieval of primary production in the sea from ocean‐color data are temperature based. But previous studies in low latitudes have shown that models that include phytoplankton community structure can have improved predictive capability. In this study, we measured photosynthetic parameters from photosynthesis‐irradiance (P‐E) experiments, phytoplankton absorption coefficients, and phytoplankton community structure derived from algal pigments during four cruises in the northern South China Sea (NSCS). The maximum quantum yield of CO2 ( ) and the chlorophyll a‐normalized P‐E curve light‐limited slope (αB) varied significantly with the blue‐to‐red ratio of phytoplankton absorption peaks (aph(435)/aph(676)) (p < 0.001, r = −0.459 and −0.332, respectively). The unexplained variability could be due in part to the absorption associated with nonphotosynthetic pigments. The chlorophyll a‐normalized light‐saturated photosynthetic rate ( ) at the surface showed a unimodal distribution over the chlorophyll a range during the spring and summer, and significantly increased when Prochlorococcus was outcompeted by other picophytoplankton (p < 0.01). Almost 60% of the variance of could be explained by a piecewise regression with phytoplankton absorption coefficients and pigment markers. Unlike previous studies, our data showed that changes of were unrelated to the size structure of phytoplankton. Although a temperature‐based approach could not effectively predict αB and in the NSCS, a trophic‐based approach can be used for assignment of these parameters in a regional primary production model using ocean‐color data.