The industry-standard surface-related multiple elimination (SRME) method provides an approximate predictor of the amplitude and phase of free-surface multiples. This approximate predictor then calls upon an energy-minimization adaptive subtraction step to bridge the difference between the SRME prediction and the actual free-surface multiple. For free-surface multiples that are proximal to other events, the criteria behind energy-minimization adaptive subtraction can be invalid. When applied under these circumstances, a proximal primary can often be damaged. To reduce the dependence on the adaptive process, a more accurate free-surface multiple prediction is required. The inverse scattering series (ISS) free-surface multiple elimination (FSME) method predicts free-surface multiples with accurate time and accurate amplitude of free-surface multiples for a multidimensional earth, directly and without any subsurface information. To quantify these differences, a comparison with analytic data was carried out, confirming that when a free-surface multiple interferes with a primary, applying SRME with adaptive subtraction can and will damage the primary, whereas ISS free-surface elimination will precisely remove the free-surface multiple without damaging the interfering primary. On the other hand, if the free-surface multiple is isolated, then SRME with adaptive subtraction can be a cost-effective toolbox choice. SRME and ISS FSME each have an important and distinct role to play in the seismic toolbox, and each method is the indicated choice under different circumstances.