Abstract: Hydro-mechanical continuously variable transmission (HMT) can often dominate the integral economic performance of tractors in sustainable agriculture. The HMT efficiency of the tractors can fluctuate greatly with the change in input and output conditions. These fluctuations can seriously affect the fuel consumption and power output of tractors. Ultimately, the operating cost of tractors has been one of the most key factors in the agricultural machinery industry with fuel consumption. This study aims to optimize the engine and HMT operating point using transmission efficiency. Two strategies of economic control were taken: the engine efficiency optimization control strategy (EOCS) with the highest engine efficiency as the objective function, and the integral efficiency optimization control strategy (IOCS) with the highest efficiency of the integral engine and HMT as the objective function. The engine and integral system efficiency were then obtained to evaluate the performance and economy of tractors under the different control strategies. Firstly, the efficiency models were established for the engine and HMT. The control parameters (engine speed and HMT speed ratio) were then optimized cooperatively for the highest operating efficiency of the HMT tractors. The improved algorithm of the fish swarm was utilized to determine the optimal speed of the engine and the speed ratio of HMT under two control strategies. The best economic matching of engine speed and HMT speed ratio was obtained under various conditions. The model of the HMT tractor was simulated using the Matlab/Simulink platform. The real data of the tractor was collected to more accurately simulate the performance of the tractor under typical working conditions. A cyclic working model was constructed to take the vehicle speed and working resistance as the key parameters during simulation under real conditions. The data segments of common working conditions were extracted from the road/operation test data of real vehicles. These working conditions were then selected to simulate the typical operational scenarios, such as plowing or hauling loads. A comparison was then made on the effectiveness of the two control strategies. The simulation results show that there was an outstanding difference in the fuel consumption of the two control strategies. The fuel consumption values of IOCS and EOCS were 1.356 and 1.381 L under the cyclic working conditions, respectively. The fuel consumption of IOCS was 1.81% lower than that of EOCS. The integral system efficiency of the engine and HMT was optimized more effectively for a better economy, with a decrease in the fuel consumption of IOCS. But there were some differences in the acceleration performance between the two strategies. Although the IOCS strategy shared the lower fuel consumption, there were more drastic variations in the engine operating point under different loads. The unstable performance of acceleration failed to meet the requirement of the environments, where the speed was changed rapidly and smoothly. In contrast, the engine operating point of the EOCS was changed more smoothly under different loads, although a relatively higher fuel consumption was observed. The performance of acceleration was better than that of IOCS, because the reserve torque was higher than that. The smoother response of acceleration was specifically beneficial in the required environment with rapid acceleration and stable power output. Overall, the optimal control strategies were selected to fully meet the specific requirements of operation. While the IOCS strategy was more economical under different loads. The EOCS strategy was more suitable for the operation when the performance of acceleration was more important. The finding can also provide a strong reference for the control strategies of HMT tractors, in order to fully meet the requirements of engineering applications.