Typical approaches to regulating sound performance of vehicles and products rely upon A-weighted sound pressure level or sound power level. It is well known that these parameters do not provide a complete picture of the customer’s perception of the product and may mislead engineering efforts for product improvement. A leading manufacturer of agricultural equipment set out to implement a process to include sound quality targets in its product engineering cycle. First, meaningful vehicle level targets were set for a tractor by conducting extensive jury evaluation testing and by using objective metrics that represent the customer’s subjective preference for sound. Sensitivity studies (“what-if” games) were then conducted, using the predicted sound quality (SQ) index as validation metric, to define the impact on the SQ performance of different noise components (frequency ranges, tones, transients). Once established, the vehicle level SQ targets were cascaded down to component level targets, these being either targets on the strength of the acoustic and structural sources (such as Sound Power and Forces) or targets on the vehicle sensitivities (such as vibration, acoustic and vibro-acoustic Transfer Functions). The target cascading was done by leveraging the experimental time-domain SPC (Source-Path-Contribution) model of the tractor, where acoustic sources were the main systems outside the cab (source 1, source 2, …, source N) and structural sources were attached to the cab. This paper describes the procedure followed, presents the results, and discusses the implications on the vehicle engineering process.