AbstractA key challenge for current‐generation Earth system models (ESMs) is the simulation of the penetration of sinking particulate organic carbon (POC) into the ocean interior, which has implications for projections of future oceanic carbon sequestration in a warming climate. This paper presents a new, cost‐efficient, mechanistic 1D model that prognostically calculates POC fluxes by carrying four component particles in two different size classes. Gravitational settling and removal/transformation processes are represented explicitly through parameterizations that incorporate the effects of particle size and density, dissolved oxygen, calcite and aragonite saturation states, and seawater temperature, density, and viscosity. The model reproduces the observed POC flux attenuation at 22 locations in the North Atlantic and North Pacific. The model is applied over a global ocean domain with seawater properties prescribed from observation‐based climatologies in order to address an important scientific question: What controls the spatial pattern of mesopelagic POC transfer efficiency? The simulated vertical POC transfer is more efficient at high latitudes than at low latitudes with the exception of oxygen minimum zones, which is consistent with recent inverse modeling and neutrally buoyant sediment trap studies. Here, model experiments show that the relative abundance of large‐sized, rapidly sinking particles and the slower rate of remineralization at high latitudes compensate for the region's lack of calcium carbonate ballast and the cold‐water viscous resistance, leading to higher transfer efficiencies compared to low‐latitude regions. The model could be deployed in ESMs in order to diagnose the impacts of climate change on oceanic carbon sequestration and vice versa.