Abstract: The farmland use transition is an inevitable requirement to adapt to the development of agricultural modernization. The synergistic mechanism between farmland use transition and eco-efficiency has become an important path to promote the sustainable development of urban-rural integration. This paper takes the national coordinated urban-rural comprehensive supporting reform pilot area as the research area, based on the panel data of Chengdu-Chongqing pilot area from 2005 to 2022, using the Sparrow Search Algorithm-based Projection Pursuit (SSA-PPM) model, super-efficiency SBM model, Panel Quantile Model, and the threshold regression model to explore the impact of farmland use transition on the farmland use eco-efficiency. The study found that: 1) From 2005 to 2022, the farmland use transition in the Chengdu-Chongqing Pilot Zone generally showed a fluctuating downward trend, with the average value decreasing from 1.633 to 1.536. Meanwhile, the farmland use eco-efficiency exhibited an overall upward trend, with the average value rising from 0.688 to 0.711. The overall eco-efficiency value remained below 1. DEA results indicate relative inefficiency, with losses in the farmland use eco-efficiency. However, the number of DEA-efficient counties showed an upward trend, increasing from 4 in 2005 to 8 in 2022. Spatially, both the farmland use transition and eco-efficiency exhibited a zoned differentiation pattern characterized by “low-to-medium efficiency in the center and high efficiency at the periphery.” High-value counties and districts are concentrated in areas rich in arable land resources surrounding the Chengdu-Chongqing Pilot Zone, such as Tongnan, Hechuan, Tongliang, Jintang, and Jianyang. 2) The farmland use transition in the Chengdu-Chongqing pilot area, along with its explicit and implicit transformations, has demonstrated a significant positive impact on eco-efficiency, with coefficients all exceeding 0.5. The transition of arable land utilization significantly enhanced eco-efficiency, with impact coefficients ranging from 0.532 to 0.585 across low to high quantiles. In the Chengdu-Chongqing pilot area and parts of Chengdu, quantitative transition, spatial structure transition, production function transition, and ecological function transition of arable farmland use all exerted positive effects on eco-efficiency, with the strongest positive impacts observed under high quantile conditions. In the Chongqing portion, the regression coefficients for spatial structure transition and production function transition monotonically decreased with increasing quantiles. However, influenced by arable land resource conditions and topography, the Chengdu-Chongqing pilot area exhibited more pronounced latent and explicit transitions in arable farmland use in both Chengdu and Chongqing. 3) The influence of farmland use transition on the farmland use eco-efficiency in the Chengdu-Chongqing pilot area is moderated by the dual threshold effect of urbanization level, with threshold values of 0.3612 and 0.7511 respectively. In Chongqing and Chengdu, as urbanization levels continue to rise, the positive impact of farmland use transition on eco-efficiency first declines and then increases. In the Chongqing part, it is moderated by the single threshold effect of financial support for agriculture, with a threshold value of 0.3983. The promotion effect of farmland use transition on eco-efficiency is nonlinearly weakening. The research findings provide a scientific basis for effectively implementing comprehensive land use management to enhance regional farmland use transition and eco-efficiency, offering scientific guidance for regional urban-rural integration and green development.