隐含碳排放限值是建筑结构碳排放量化调控与减碳目标保障的关键指标。本文从社会“碳中和”所需减碳路径出发，依据建筑结构保有与需求量中的新建与既有结构构成比例，提出减碳目标分解方法，为既有结构低碳维护与新建结构低碳设计提供与宏观年度减碳需求相匹配的限值预设依据。依托“双碳”目标的减碳需求构建行业预期发展情景，得出在维持现状、常规预估、拆除限制、减量预估四类典型情景下，2022年我国新建建筑结构隐含碳排放限值分别为442.6 kg CO₂e·m^-2、456.2 kg CO₂e·m^-2、485.9 kg CO₂e·m^-2、616.0 kg CO₂e·m^-2（对应结构设计使用年限50年）；给出2022—2060年建筑结构建造与维护碳排放限值的变化趋势，厘清了新建结构总量控制、既有结构延寿等减碳措施促使结构单体隐含碳排放限值宽松的成效。进一步从概率化调控、区域特征量化、可操作性、数据驱动优化等方面，提出了建筑结构隐含碳排放限值设定的相关建议。

The embodied carbon emission limit is a key parameter for quantitatively regulating carbon emissions from building structures and facilitating the achievement of carbon reduction targets. Starting from the social carbon emission reduction pathway toward carbon neutrality, a carbon emission reduction target decomposition approach is proposed based on the proportions of newly built and existing buildings in building structure stock and demand, which provides limit presetting references for existing structure maintenance and new structure design. Then, the expected development scenarios of building construction in China are established referring to the carbon peaking and neutrality goals in China, and correspondingly, the carbon emission limits for four typical scenarios, i.e., business as usual, conventional expectation, regulating demolition, and reduction expectation, are respectively evaluated to be 442.6 kg CO₂e·m^-2, 456.2 kg CO₂e·m^-2, 485.9 kg CO₂e·m^-2, and 616.0 kg CO₂e·m^-2 (the designed service life is 50 years). Besides, the expected trends of carbon emission limits during the year 2022 and 2060 have also been estimated, and the contributions of emission reduction measures, including the amount regulation of newly built structures and the lifespan extension of existing structures, to the relaxation of carbon emission limits for individual structures have been clarified. Furthermore, suggestions for research related to carbon emission limit are provided from the aspects of probabilistic regulation, quantification of regional characteristics, feasibility of the target achievement, and data-driven refinement.