Cropland abandonment exerts far-reaching consequences for food security and sustainable agricultural development. However, while it is generally believed that low agricultural comparative return (ACR) is the main cause of cropland abandonment, the magnitude of such effects and their underlying mechanisms are not well understood. On the basis of constructing a mathematical model and analysis framework, this chapter employs a large-scale farm-level panel data from Chinese Family Databases during 2015–2021 and a panel high-dimensional fixed effects model to investigate the impact of ACR on cropland abandonment. Results show that ACR has a positive impact on the proportion of abandoned cropland area to total cropland area, with an average coefficient of 0.016. This effect is mainly achieved through reductions in long-term land investment, agricultural capital input, and agricultural labor input. Nevertheless, it is also moderated by the effects of rural factor markets and government agricultural subsidies. In particular, the agricultural machinery market, the land rental market and the level of agricultural subsidies can weaken ACR’s negative impact on land use, while current agricultural subsidies have little effect. Our findings also reveal that the effects of ACR on cropland abandonment vary significantly across geographical regions, grain areas, and terrain zones. Specifically, ACR is not harmful to households in northeastern and central China or major grain producing areas, while it harms others. Overall, this study provides a better understanding of the quantitative relationship between ACR and cropland abandonment in rural China, and valuable insights for other developing countries in formulating targeted measures to curb cropland abandonment in low ACR situations.