The sudden global outbreak of coronavirus diseases 2019 (COVID-19) has infected over seventy million people and resulted in over one million deaths by the end of 2020, posing a significant threat to human health. As potential carries of the novel coronavirus, exhaled airflow of infected individuals via coughs, are significant in virus transmission. This study measures human coughs' airflow velocity in a chamber filled with stage fog employing a particle image velocimetry (PIV) system. The purpose of this study is to examine and provide accurate boundary conditions for the prediction of the virus transmission routes using computational fluid dynamics (CFD) simulations. Sixty cough cases from ten healthy nonsmoking volunteers (five male and five female, averaged age of 29.3±4.0) are taken respectively, and ensemble-average operations are conducted to eliminate individual variations. Velocity distribution measurements are obtained in the vertical and horizontal planes around the mouth area. Temporal and spatial cough flow ensemble-averaged velocity profiles and standard deviations, cough duration time (CDT), peak velocity time (PVT), maximum cough velocities, and average spread angle of the cough jet are measured. Results show that the CDT of the cough airflow is 520–560 ms, and PVT is 20 ms. The male/ female averaged maximum velocity is 15.2/13.1 m/s, respectively. The average vertical/horizontal spread angle from the mouth is 15.3°/13.3° for males and 15.6°/14.2° for females, respectively. With the measurement data, it is possible to refine the initial boundary conditions of a simulated cough and model cough flows more accurately.