A vortex whistle is a low-cost but accurate flow meter. As the airflow passes through the whistle, a tone is produced which has a frequency proportional to the inlet air flow rate. Recent investigations using vortex whistles have focused on the use of this relationship to quantify aspects of respiratory function. Despite promising results, there is a lack of understanding of the physical mechanisms underlying vortex whistle function. In the present study, combined CFD and CAA method is applied to investigate the performance of the vortex whistle. A simplified vortex whistle model is simulated by LES with Smagorinskty model to predict the flow field within and outside the whistle, and then Ffocus-Williams and Hawkings analogy is used for aeroacoustic simulation to predict the acoustic response of the vortex whistle and the frequency and level of the signature tones. The result indicates that the spectral peaks are generated by the cylindrical cavity of the vortex whistle. The whistles with different heights and diameters of cylindrical cavity are also simulated to optimize the design. The results show that the height does not affect the harmonic frequency significantly; with the decrease of the diameter, the harmonic shifts to a higher frequency. It is beneficial for low flow rate cases; in addition, an optimal design exists which presents the largest signal-to-noise ratio. The simulation results and analysis in the present study help us determine which part contributes the signature peak, and increased SNR of the acoustic output by design optimization will allow for improved frequency-to-flow mapping and volume estimates during respiratory tasks such as the elicitation of vital capacity.