Abstract Based on the manufacturing processes panel data for 42 iron and steel enterprises (ISEs) in China from 2005 to 2014, a data envelopment analysis (DEA) model with undesirable output and a Malmquist-Luenberger (ML) index model are used to calculate the pollution control efficiency (PCE) and its variation trend of China's 42 ISEs. The results show that the comprehensive PCE of the 42 ISEs is relatively low, and the PCE of short-process enterprises is better than that of long-process enterprises. The total factor pollution control efficiency (TFPCE) of China's ISEs with a short process presents an increasing trend, which is mainly attributed to the increase of the technical change, while the efficiency change has a slight negative effect. In contrast, the TFPCE of ISEs with a long process is in the trend of decreasing, which is mainly affected by the decline of the technical change, while the efficiency change has a slight positive effect. In addition, the PCE varies significantly across regions and ISEs. Among 42 ISEs, the TFPCE of 18 enterprises shows an increasing trend, while that of 24 enterprises presents a decreasing trend.