Agricultural carbon productivity combines the dual attributes of reducing carbon emissions and stabilizing economic growth, and is a core aspect of the new era of low-carbon agricultural development. The construction of high-standard farmland is an important initiative to promote high-yield and high-efficiency agriculture, as well as environmentally sustainable development through land improvement in China. However, the impact of high-standard farmland construction on agricultural carbon productivity and its mechanisms is still in the process of being determined. In order to fill this gap, this study aims to construct a theoretical, analytical framework for the impact of high-standard farmland construction on agricultural carbon productivity. Based on the current situation of high-standard farmland construction and the characteristics of agricultural carbon productivity in China, this study used the panel data of 31 provinces, from 2003 to 2017 in China, to empirically examine the effects, mediating paths, and spatial characteristics of the construction of high-standard farmland on agricultural carbon productivity using a double fixed-effects regression model, a mediating-effects model, and a spatial econometric model. The results show that: (1) High-standard farmland construction has a positive effect on agricultural carbon productivity, with a direct effect coefficient of 0.139 after adding a series of control variables. (2) Furthermore, heterogeneity analysis shows that the impact of high-standard farmland construction on agricultural carbon productivity will vary greatly depending on the topographic characteristics of the studied area, the level of economic development, and whether it is a main grain-producing area. (3) Mechanism analysis shows that agricultural scale operation, agricultural planting structure, and agricultural technology progress all have partial mediating roles in the impact of high-standard farmland construction on agricultural carbon productivity, with mediating effect coefficients of 0.025, 0.024, and 0.013, respectively. (4) Agricultural carbon productivity has a spatial correlation, and for every 1% increase in the level of high-standard farmland construction, agricultural carbon productivity increases by 0.117%, with a direct effect of 0.074% and a spatial spillover effect of 0.043%. Our study explains the impact effects, mechanisms, and spatial spillover effects of high-standard farmland construction on agricultural carbon productivity from theoretical and empirical perspectives, thus deepening the literature on the relationship between high-standard farmland construction and agricultural carbon productivity, and providing a theoretical basis and practical references for improving agricultural carbon productivity from the perspective of high-standard farmland construction policy.