Beef calf production is a source of sustainable development for the beef cattle industry. However, no comparative studies have reported on the technical efficiency of different beef calf production systems and their influencing factors. Based on data from 218 Chinese farmers and 12 governments, in the present study, we constructed data envelopment analysis (DEA) models and conducted a comparative analysis of the technical efficiency of the main three beef calf production systems: the Simmental calf intensive production system (CIPS), Simmental calf semi-intensive production system (SCIPS) and Holstein bull calf intensive production system (BCIPS). Using Tobit models, we analyzed the effects of various factors. The results show that: (1) The technical efficiency of the production system of Simmental calf is higher than that of Holstein bull calf, and the efficiency of SCIPS is higher than that of CIPS. The technical efficiency (TE), pure technical efficiency (PTE) and scale efficiency (SE) of different systems are significantly different. (2) Policy on the environment positively affected (p < 0.01) the TE, TPE and SE of CIPS, but negatively affected the PTE of SCIPS. Therefore, appropriate environmental regulations have a positive effect on production efficiency, which means that measures should be taken according to the reality and characteristics of the production system, and policies applicable to other systems or regions may not be applicable in a given case. (3) The management level and technology training had positive effects on the TE, TPE and SE of the three systems, while the number of years of production had a negative or no significant effect. Producers are not the “perfectly rational economic man”, and the more knowledge they have, the more productive they will be. However, the “knowledge” referred to here is that which is adapted to production, not that which is traditional. The knowledge possessed by the producer should be updated continuously with the changes over time and the development of the industry, while outdated information is not considered as “knowledge” here. Therefore, to achieve sustainability, the government should fully consider the characteristics of the local breeding mode and, more importantly, the expected effects of policies to be implemented.