This work presents an experimental and analytical investigation of the distortion of grid generated turbulence and its relation to the aeroacoustics of a low Reynolds number propeller. The propeller has a chord based Reynolds number of 10⁴≤ Re_c  ≤ 10⁵ and operates at an advance ratio of 0.225. The mean flow model of Hough and Ordway (H&O) is used to compute the streamtube upstream of the propeller. This allows the distortion of turbulence due to streamtube contraction to be predicted by means of Paterson and Amiet's distortion model. The distortion of turbulence by the leading edge of the blade is modelled with the asymptotic solution of Hunt. Amiet's propeller turbulence ingestion noise model is used to couple the distorted turbulence to noise. The models are validated by means of an experiment. A square bar grid with mesh size 0.1m and bar width 0.01m is placed upstream to generate turbulence. A load-torque cell measures the thrust which serves as input for the analytical mean flow model. Using particle image velocimetry (PIV) and hotwire anemometry the turbulence statistics upstream and up to the propeller plane are measured. PIV is also used to measure the distortion of turbulence by the leading edge of the propeller blade. The experiment is done in the TU Delft A-tunnel, where acoustic measurements are performed using a microphone array. Results reveal that the model of H&O overpredicts the induced velocities and a correction has to be applied such that the experimental results are matched. The analytical model and flow measurements show that the isotropic turbulence is not distorted due to the contraction of the streamtube and isotropy is conserved. The turbulence intensity and integral lengthscale do change and this is ascribed to the decay of turbulence behind a grid. Turbulence is distorted by the leading edge and an increase of the turbulent upwash velocity of 25% is observed. It is found that by using the decayed turbulence statistics, the tones at the blade passing frequency harmonics are best resembled. The correction for distortion by the leading edge does not improve the predictions.