Summary Pre-stack data interpolation is always an attractive issue for the potential related acquisition cost reductions. Once designed on the base of an acceptable cost-quality trade-off, a field layout can be adequately decimated thus leaving only those shotpoints that are strictly necessary to recover the final image quality by shot-infilling. This decimated layout is called minimal acquisition (MA) layout. Of course the MA layout depends on the geophysical targets and the data quality, which are given, as well on the interpolation method and strategy, which therefore must be carefully selected. In this paper we consider the MA design based on the 3D shot continuation operator (3D SCO) infilling, mainly because 3D SCO is a pre-stack interpolation method which is specifically suitable for infilling irregular land acquisition geometries. The MA design for 3D SCO infilling is validated by decimating a full 3D land survey to different MA layouts. Results show that different decimation ratios can be achieved according to the different geophysical requirements (e.g. post-stack imaging, PreSDM, AVO, etc.). The quality of infilled data is strongly dependent not only on the quantity of deleted shots, but also on their location. The optimization of the shot locations is key in the MA design and it calls for the definition of new design criteria to control the decimation of the planned acquisition by 3D SCO. The methodology 3D land acquisition costs can be reduced by removing, from the optimized acquisition geometry, those shotpoints that can be interpolated from the outlasted shots without affecting the required final data quality at the target. The reduced configuration, here referred to as minimal acquisition (MA) layout, depends mainly on the interpolation methodology. Each interpolation algorithm has its own limitations and drawbacks that must be taken into account when relaxing the acquisition geometries. The 3D shot continuation operator (3D SCO) is a model-based algorithm for wavefield interpolation of pre-stack data that is based on a Kirchhoff-type implementation [1]. Its main benefit when dealing with 3D land surveys is the capability to infill irregular acquisition geometries to arbitrary (regular or irregular) geometries. Therefore, 3D SCO is better suited to reconstruct missing shots, rather than other interpolation methods that require regular grids [3]. However, 3D SCO requires a different approach to MA layout design, as the quality of infilled data depends on the coverage of the operator itself, which is not adequately described by indicators commonly used in acquisition design, such as fold of coverage. Another aspect to be taken into account is the accuracy in the velocity model estimation, since 3D SCO is modelbased. It can be shown that insensitivity to errors in the velocity model is strictly dependent on the interpolation strategy: the optimal strategy should result from a trade off between operator coverage requirements and minimization of continuation distance [2]. Suitable indicators can