Statistics can guide the direction of reliable alternatives in minimizing casualties caused by marine accidents, deriving preventive factors to prevent accidents in advance, and further discovering elements to improve the response capabilities of rescue authorities. In addition, statistics on marine accidents and Search and Rescue operations can be a vital indicator in measuring the cost-effectiveness of the national budget spent on preventing, preparing, and responding to accidents and serve as a monitoring window for the public to assess their level of marine safety. This study compared analyzed the marine accident statistics system operated by the Korea Coast Guard with the US Coast Guard’s statistics system, which is recognized as an outstanding model for search and rescue statistics management, to derive valuable implications for reforming the framework of Korean search and rescue statistics from ‘occurrence-oriented’ to ‘response -oriented.’ As a result, measures that can be improved in the short term, such as fractionating sub-categories of the statistical system’s input data, adjusting unrealistic formulas for lifesaving rates on SAR cases, and setting the input criteria as usable accident data to secure statistical reliability, were suggested. Furthermore, from a mid-to-long-term perspective, designing a state-of-art search and rescue statistics management platform based on big data was proposed.