The self-assembly of copolymeric vesicles and micelles in micromixers is studied by External Potential Dynamics (EPD) simulations - a dynamic density functional approach that explicitly accounts for the polymer architecture both at the level of thermodynamics and dynamics. Specifically, we focus on the co-solvent method, where nanoparticle precipitation is triggered by mixing a poor co-solvent into a homogeneous copolymer solution in a micromixer. Experimentally, it has been reported that the flow rate in the micromixers influences the size of the resulting particles as well as their morphology: At small flow rates, vesicles dominate; with increasing flow rate, more and more micelles form, and the size of the particles decreases. Our simulation model is based on the assumption that the flow rate mainly sets the rate of mixing of solvent and co-solvent. The simulations reproduce the experimental observations at an almost quantitative level and provide insight into the underlying physical mechanisms: First, they confirm an earlier conjecture according to which the size control takes place in the earliest stage of the particle self-assembly, during the spinodal decomposition of polymers and solvent. Second, they reveal a crossover between different morphological regimes as a function of mixing rate. Hence they demonstrate that varying the mixing rate in a co-solvent setup is an effective way to control two key properties of drug delivery systems, their mean size and their morphology. (c) 2017 Elsevier Ltd. All rights reserved.