Microreactors are commonly regarded by their high mass transfer rates associated to their small diameters. Although this is true, especially when compared to batch systems, mixing is frequently exclusively dominated by diffusion in single phase flow microreactors, being the fast mixing of reactants an aspect often overlooked during their design. This paper presents the first quantitative analysis of the effect of mixing of precursors on the size and distribution of metal nanoparticles synthesized in flow microreactors. Silver nanoparticles with a range of sizes are continuously synthesized by controlling mixing efficiency in rationally-designed microreactors. The particle size decreases as the mixing index increases due to an increase in nucleation rate and thus the resulting nuclei concentration. Herein, the mixing efficiencies of different 3D curved microreactors are quantitatively evaluated using accurate simulations of their concentration profiles, avoiding the so-called numerical diffusion errors using a novel method based on backward particle tracking. A method based on forward particle tracking is improved to simulate the residence time distributions (RTD) in microreactors at a reduced computational cost. Curving the reactors channels is known to lead to the formation of Dean vortices however, these well-defined rotations lead to stagnant zones. The mixing efficiency can be enhanced and the RTD narrowed by periodically changing the direction of the Dean vortices along the length of the channel. In addition, we demonstrate the effect of the configuration of the inlet streams relative to the curvature of the reactor and thus, the Dean vortices, on the resulting mixing. These results provide rational design guidelines for the design of microreactors to manipulate advection as a way of manipulating the reaction rates as demonstrated here to control nanoparticle sizes.