Seismic imaging beneath shallow gas is a challenge that needs to be addressed through a multifaceted solution involving advances in acquisition, pre-processing, and depth imaging. The challenge results from the host of geophysical issues that are incurred as seismic energy is incident on gas in the near surface. Energy absorption by shallow gas leads to seismic attenuation of high-frequencies and necessitates a broadband acquisition and processing solution to ensure the presence of signal at the reservoir level. This absorption and internal scattering present challenges to conventional velocity model building techniques that are dependent on reflection traveltime information, due to the lack of strong, coherent reflections in common image point (CIP) gathers. However it is essential for these velocities to be properly delineated. A velocity model that does not resolve these shallow pockets of gas-saturated sediments can create non-geological structural undulations in depth-migrated volumes, impacting on the volumetric estimated at the reservoir level. In addition they present the risk of misinterpretation and of unprepared drilling through over-pressured zones. To mitigate this problem, a two-way wave equation approach was employed to derive a velocity model via prestack acoustic full-waveform inversion (FWI).FWI is a waveform-based velocity model building algorithm that operates in the data domain. It is a semi-automated process to obtain a high-resolution earth model that best explains the amplitude and phase information of the acquired data by iteratively minimizing the least-squares difference between acquired data and forward-modelled data. This technology has emerged over the last decade from being an academic exercise to a readily available pragmatic depth-imaging tool, with significant contributions to petroleum exploration. These contributions have occurred in a range of geological settings throughout the world, from deep-water salt basins in the Gulf of Mexico to shallow-water, shallow-gas sediments in South-East Asia.This paper focuses on the concept and results of a collaborative test that was undertaken to examine how applying 3D acoustic anisotropic FWI to an optimized over/under (sparse under) broadband dataset could improve the resolution and imaging through shallow gas in a shallow-water environment.