The feasibility of aeroacoustic noise predictions based on Particle Image Velocimetry (PIV) measurements is studied. For this purpose, experiments are conducted on a sharp trailing edge (TE) flow developed along a flat plate at free stream velocity of 15m/s. The acoustic emissions were characterized in the NLR Small Anechoic Wind Tunnel (KAT) by means of microphone measurements. The result is used for benchmarking the PIV based noise predictions. PIV measurements were carried in a low-speed wind tunnel of TU Delft with similar properties to that of the KAT facility. Planar PIV measurements, performed at high spatial resolution, characterize the turbulent properties of the boundary layer travelling across the trailing edge. Time resolved Tomographic PIV measurements are obtained over a volume of 19.7mm x 9.1mm x 33.4mm at an acquisition rate of 12kHz. Two different approaches are compared for the prediction of TE noise based on PIV data. The first follows diffraction theory and requires the deduction of the instantaneous surface pressure fluctuations from PIV data. The second one relies on the integral solution of the Lighthill equation obtained through a tailored Green’s function for a semi-infinite half plane. This allows to calculate the scattered noise directly from the velocity data. The predictions based on diffraction theory agree well with the measured far field spectrum up to 4kHz. In contrast, the Green’s function approach yields a significant and systematic overprediction of approximately 12dB.