Lung sound is commonly analyzed when testing for lung abnormalities. However, the accuracy of this analysis is low and more information on sound generation and alteration is needed to improve the analysis quality. In the current study, an aeroacoustic investigation of sound generation in an airway model was performed experimentally to uncover the factors that change the sound characteristics due to bronchoconstriction. A Tbranch configuration was used to represent an airway junction, and a constricted tube in the mother branch was used to represent a bronchoconstriction. Aeroacoustic sound was analyzed at several flow rates, representing different speed maneuvers. Constriction percentages of 25%, 50%, and 75% simulated different bronchoconstriction severities. The power spectral density of the produced sound increased over a wide frequency range as the flow rate and constriction level increased. The overall sound pressure level (OASPL) over several frequency bands was calculated and it was found to be related to the Reynolds number in the smallest cross-sectional area of the constriction. When constriction was less than 50%, the OASPL in the frequency range of 200-800 Hz increased as the Reynolds number increased. In the 75% constriction case, a smaller increase of OASPL was observed. In the frequency range of 150-10 000 Hz, all models demonstrated similar relationships between OASPL and Reynolds number. In the majority of frequency ranges, a Reynolds number of 4000 was required to generate 2 dB OASPL, and OASPL showed dramatic increases with higher Reynolds numbers. To find the source location based on Lighthill's sound analogy, turbulence strength measurements were performed 5 mm downstream from the constricted area. Small turbulence was observed, indicating that the sound sources were nearby. Our results show that the OASPL increase of the lung sound can be an indicator of the constriction presents in the airway.