An acoustic analysis of a ducted fan unmanned aerial vehicle (UAV) was conducted to identify the primary aeroacoustic sources and to determine the potential for reducing the radiated noise levels. Computational fluid dynamics was used to determine the three-dimensional flow field through the ducted fan in hover and maneuvering configurations. The flow solutions provided information about the blade relative velocities, spatially non-uniform inflow, inflow turbulence, boundary layer turbulence, and blade wake velocity deficits for use in acoustical models of broadband and blade passing frequency noise. The computational results were in good agreement with experimentally measured noise levels, and showed that the tonal noise was produced primarily by unsteady forces resulting from the non-uniform inflow, while the broadband noise resulted from the inflow turbulence. Based on these findings, design modifications were recommended which offer the potential to reduce the noise by more than 10 dB.