Pressure-swirl nozzles are widely used in both industrial and experimental processes. However, the spray flow fields are difficult to predict due to the complexities of the dense break-up region. Herein, an inflow boundary condition approach is employed whereby droplets are computationally injected at 9 mm downstream of the orifice based on experimentally derived descriptions of the mean droplet and gas velocities, combined with empirical relations for turbulent kinetic energy and dissipation. The gas flow is then predicted with a Reynolds-Averaged Navier-Stokes solution while the particle trajectories are computed with a discrete random walk model. This approach was found to reasonably describe downstream spatial distribution of gas and droplet velocity, with moderate mesh resolution and CPU time. In addition, the evolution of the droplet velocity fluctuations was found to be consistent with turbulent diffusion theory based on the local Stokes numbers and turbulent kinetic energy. (C) 2016 Elsevier B.V. All rights reserved.