To reduce the water noise induced by flow over broad-crested weirs and provide a basis for engineering optimization, a combined approach integrating physical model experiments and numerical simulations was adopted to investigate the influencing factors and spectral characteristics of the water noise. Physical model experiments were conducted to examine the effects of upstream discharge, downstream water depth, and weir height on the sound pressure level of water noise and the root mean square of fluctuating pressure. Quantitative relationships between the water noise sound pressure level and these influencing factors were established, and the A-weighted spectral characteristics of the water noise were analyzed. Numerical simulations were employed to investigate the effects of droplets and bubble clusters on water noise and reveal the causes of variations in the A-weighted spectrum. The results show that both the sound pressure level of water noise and the root mean square of fluctuating pressure are positively correlated with upstream discharge and weir height, and negatively correlated with downstream water depth. Among these factors, upstream discharge exhibits the most significant influence, followed by downstream water depth and weir height. High-frequency noise generated by bubble bursting leads to the dominant frequency of the A-weighted spectrum of water noise being located in the high-frequency range. Increasing the upstream discharge and weir height raises the high-frequency sound pressure level in the A-weighted spectrum, whereas increasing the downstream water depth reduces the high-frequency sound pressure level.