Heat stress caused by climate change and heat-related labor productivity losses have become global concerns. Estimating the economic impacts of heat stress is of great significance for employers, as well as sectoral and national policy makers who are searching for solutions to reduce productivity losses. As the value of economic impacts are sensitive to the research methodologies, we conducted a systematic review of published literature on the methodologies and results of economic impacts of heat on labor productivity. Four methods were summarized: the human capital (HC) method, the econometric method (EM), the input–output (IO) method, and the computable general equilibrium (CGE) model. Considering adaptation measures, global economic losses due to heat-related labor productivity losses are projected to range from 0.31% (0.14–0.5%, RCP2.6) to 2.6% (1.4–4%, RCP8.5) of global GDP in 2100. The published studies found that large economic losses occurred mainly in South and Southeast Asia, Sub-Saharan Africa, and Central America. Owing to different methodologies and considerations of adaptation measures, the disparities of results within the same area at a given time can be as high as 7.4-fold. We summarized the knowledge gaps in existing studies and proposed new directions to provide more targeted and reliable results for policy makers.