Abstract Fourier-based Nearfield Acoustical Holography (NAH), Statistically Optimized Nearfield Acoustical Holography (SONAH) and the Equivalent Source Method (ESM) are widely used in noise source identification and as important tools to guide design modification for noise control purposes. Fourier transform-based NAH requires the sound field to fall to negligible levels outside the measurement aperture, which is a requirement that is rarely met in practice. To overcome this difficulty, SONAH and ESM have been developed. In addition, the Inverse Boundary Element Method (IBEM) can also be used, given sufficient computational resources. Unfortunately, none of these methods can directly guide the design modifications required to unequivocally reduce noise radiation from sources. Previously, radiation mode analysis has been primarily associated with the forward prediction of sound power radiated from noise sources. Since radiation modes contribute independently to the sound power radiation, it is only necessary to modify the surface vibration so that it is not strongly coupled with those modes having high radiation efficiencies in order to ensure sound power reduction. In the current work, an inverse method based on radiation modes was investigated, in which the radiation modes were used as the basis functions to describe surface motion of a source. Thus, this procedure allows the surface vibration that results in the majority of the radiated sound power to be identified unequivocally, and so will, in turn, guide the design changes needed to reduce radiated sound power.