The sound radiated during a tennis impact has been shown to greatly influence the perception of 'feel' but has also been suggested to contribute to the perception of equipment quality, as is the case with many other consumer products. In an industry that is governed by design limitations, tennis racket manufactures are continually aiming to differentiate their products from their competitors' and the sound character of a racket is seen as one such method to do so. In order to control the sound radiated from a tennis racket it was first necessary to identify how a tennis racket radiates sound. Through a number of controlled player tests, involving groundstrokes and serves, the sound was recorded along with measurements of the physical vibrations excited in the frame. Analysis of the data revealed typical characteristics of the sound in the time and frequency domain. The sound was split into two sections to aid the analysis; an initial impulsive component that decayed very quickly and a ringing component that was of much lower amplitude but decayed at a much slower rate. The evolution of the frequency content over time was also investigated, however, the data provided much more information as to where each frequency component originated by analysing the data together with the experimental modal analysis data. The experimental modal analysis of a tennis racket was a vital stage in understanding which components of the racket were responsible for radiating the identified frequency component in the sound spectrum. The investigations identified frame bending modes, out-of-plane and inplane, as well as torsional modes, hoop modes and stringbed modes. To enable direct comparison between the natural frequencies excited during a tennis shot and the frequencies recorded from a freely suspended racket, the effect of the hand on the modal behaviour of the racket was analysed; experimental modal analysis data from a hand-gripped racket was compared with data generated by adding simulated mass to the modal model of the freely suspended racket. The first stringbed mode was identified as a key contributor to the sound of the racket, especially in the ringing component of the sound. Analysis of the physical vibrations in the frame following a tennis shot revealed that the first stringbed mode excited the frame of the racket, which is iv thought to be the reason why stringbed modes contribute to the sound more so than their surface area would suggest that they are capable of. Analysis of the relationship between subjective perceptions of players and calculated sound metrics suggest a negative correlation exists between the duration and loudness of a tennis shot and the appeal of the sound. Further investigations involving a jury to evaluate the appeal of a tennis impact sound, modified in terms of duration, revealed further evidence to support the theory that tennis impact sounds of greater duration are less appealing than those that decay quickly. This is the first study to investigate the sound radiated from a tennis racket and has identified how the sound is radiated from the racket and also suggests which parameters of the sound are considered appealing to tennis players. With this information it is possible to suggest design modifications that would influence the sound of a tennis racket in a predictable manner.