Abstract:This paper applies SE-DEA window model to value the input and output efficiency of marine science and technology of 11 coastal provinces and cities in China from 2006 to 2012. Then, it applies the Malmquist index analysis and clustering analysis to explore the regional disparity and the reasons of efficiency change. The research finds: at the national level, the overall growth of efficiency is mainly caused by the improvement of management level and scale optimization of marine science and technology activities; at regional level, regional efficiency changes can be divided into three types, accumulation-driven, application-driven, and low-growth. The main factors driving the efficiency of marine science and technology growth varies by region, however, the decrease of efficiency is mainly due to the lack of accumulation and improvement of marine scientific and technological knowledge.