A fundamental ecological trait of marine organisms is body size. Various theoretical models have used the size distribution of plankton communities to explain their trophic structure and functioning. Recent studies indicate that changes associated with global warming, eutrophication, and fisheries might have dramatically shifted the size structure of marine organisms and thus changed ecosystem functioning. Accordingly, size structure has been suggested as a useful indicator for monitoring ecosystem status. Here, we used three size metrics to represent mesozooplankton size structure: the slope of normalized biomass spectra (NB-SS), NB-SS linear fit, and size diversity. We analyzed the relationship between zooplankton size structure and environmental conditions in the very dynamic East China Sea (ECS). We tested four hypotheses predicted by theoretical studies: (1) a coastal–offshore gradient for NB-SS, (2) a negative relationship between temperature and NB-SS steepness, (3) a positive relationship between trophic status (i.e., chlorophyll a and inorganic-nutrient concentrations as a proxy) and NB-SS steepness, and (4) a positive relationship between oligotrophy and NB-SS linearity. Hypotheses 1–3 did not stand up to our analyses of the empirical data, while hypothesis 4 was supported. In the ECS, we found no clear spatial gradient for each of the three size metrics, where (i) size diversity was not correlated to temperature and significantly decreased with chl a and phosphate concentrations, (ii) the NB-SS slope showed no relationship with the environment, and (iii) NB-SS linearity tended to be higher in more oligotrophic offshore waters. Trophic status exerted a stronger effect than temperature on size distribution, but its effect was opposite to model prediction. Our results suggest that population dynamics (i.e., pulses of reproduction due to increased food availability) override metabolic scaling effects on mesozooplankton NB-SS at the mesoscale. In addition, the estimation of NB-SS slopes depends on NB-SS linearity and may therefore be less accurate than size diversity as a synthetic indicator of NB-SS changes. Our results highlight the need for caution when using size-based indicators for monitoring environmental forcing on communities in highly dynamic ecosystems.