Aeroacoustic whistles are a significant source of customer complaints to automotive manufacturers. Whistles can occur on many such components, but the relative position and configuration of rearview mirrors means they are a more problematic source of tonal noise on vehicles. The low subsonic complex turbulent flow, combined with small cavity scales, determines the possible whistle mechanisms. The one considered to be most problematic, fluid-dynamic cavity resonance, is the topic of this research thesis. The research scope is limited to the automotive environment of external rearview mirrors and the fluid-dynamic resonance mechanism: low subsonic Mach number, M = 0.05 - 0.13; laminar boundary layers; and two-dimensional, acoustically compact cavities. The low unit-cost of rearview mirrors and the desire to have simple identification and prediction schemes, that could be used by production engineers, determined an empirical approach. A search of the existing literature revealed that there were some data on cavities of the above scale in low Mach number flow, but quoted errors in empirical descriptions were large and there was very little research on the effects of flow yaw angle on the chosen resonance mechanism. The research therefore aims to determine whether existing empirical descriptions of fluid-dynamic cavity resonance are suitable for the prediction of the resonance characteristics, with sufficient accuracy to enable unambiguous identification of the presence of the resonance and its mechanism. A second aim is to investigate the effects of a feature of the automotive flow environment, flow yaw angle, on the resonance. Flow yaw angle is determined by those components of the flow in the same plane as the surface in which the cavity is situated. An experimental program was undertaken using a purpose-built aeroacoustic wind tunnel and a simple cavity model. Testing with two types of cavity configurations, as well as flow visualisation, investigated the main features of the resonance in time-averaged yawed flow. Within the scope of this thesis, it is shown that fluid-dynamic cavity resonance characteristics can be accurately identified by a simple empirical model, even in yawed flow. Various descriptors allow identification of the resonance threshold, stage, frequency and relative amplitude in non-yawed flow, while the frequency and stage can also be identified in yawed flow. The relative decrease in resonance amplitude in yawed flow, although identified for these experiments, would depend on the degree of spanwise variation in the boundary layer characteristics for a given cavity configuration. The results also identify significant issues with testing in a free jet tunnel, due to the nature of fluid-dynamic cavity resonance and the fluctuation energy content in free shear layers. Despite this, the thesis aims are achieved, and appropriate design guidelines are produced for automotive designers.