Abstract: Smart agriculture is a new agricultural application model that deeply integrates the new generation of artificial intelligence technologies such as deep learning, machine vision, knowledge engineering, reasoning and decision making with agricultural scenarios. Tobacco planting is mainly distributed in hilly and mountainous areas in China, and tobacco intelligent production showed both the common characteristics and certain specificity compared with agricultural intelligent production of majior crops. To systematically review the research and practice status, and to analyze the technical bottlenecks restricting the development and the developing trends of intelligent tobacco production in China, intelligent perception, intelligent analysis and intelligent decision-making were analyzed in key processes such as seedling, field, flue-curing, grading and alcoholization in this study. And problems such as intelligent sensing special equipments, data integration, key algorithms and multi-functional decision-making platform which need to be paid attention to in the development of tobacco smart production in China and suggestions for further development were put forward. The results in this study can provide references for the research and practice of China's intelligent tobacco production.