An ecosystem model was used to (1) determine the extent to which global trends in the ratio of mesozooplankton production to primary production (referred to herein as the " z-ratio") can be explained by nutrient enrichment, temperature, and euphotic zone depth, and (2) quantitatively diagnose the mechanisms driving these trends. Equilibrium model solutions were calibrated to observed and empirically derived patterns in phytoplankton biomass and growth rates, mesozooplankton biomass and growth rates, and the fraction of phytoplankton that are large (>5 μm ESD). This constrained several otherwise highly uncertain model parameters. Most notably, half-saturation constants for zooplankton feeding were constrained by the biomass and growth rates of their prey populations, and low zooplankton basal metabolic rates were required to match observations from oligotrophic ecosystems. Calibrated model solutions had no major biases and produced median z-ratios and ranges consistent with estimates. However, much of the variability around the median values in the calibration dataset (72 points) could not be explained. Model results were then compared with an extended global compilation of z-ratio estimates (>10 000 points). This revealed a modest yet significant ( r=0.40) increasing trend in z-ratios from values ∼0.01-0.04 to ∼0.1-0.2 with increasing primary productivity, with the transition from low to high z-ratios occurring at lower primary productivity in cold-water ecosystems. Two mechanisms, both linked to increasing phytoplankton biomass, were responsible: (1) zooplankton gross growth efficiencies increased as their ingestion rates became much greater than basal metabolic rates and (2) the trophic distance between primary producers and mesozooplankton shortened as primary production shifted toward large phytoplankton. Mechanism (1) was most important during the transition from low to moderate productivity ecosystems and mechanism (2) was responsible for a relatively abrupt transition to values >0.1 in high productivity ecosystems. Substantial z-ratio variations overlying these mean trends remained unexplained by these mechanisms. Potential sources of this variability include zooplankton patchiness, unresolved effects of advection and unsteady dynamics, unresolved shifts in mesozooplankton sizes and species, and unresolved aspects of zooplankton bioenergetics. Comparison of the modeled z-ratio patterns and mechanisms diagnosed herein with those obtained using models with expanded biological dynamics embedded in global circulation models will help further elucidate the causes of this variation.