Global warming produced mainly by the emission of greenhouse gases is currently a worldwide concern. In the last few decades since the 1950s many of the changes observed in the world climate have been meaningful. This paper presents an analysis of energy efficiency and of greenhouse gas emissions in the Chilean manufacturing industry by region and sector taking into consideration time sequences. Data Envelopment Analysis (DEA) models are used for the analysis. Three ways to handle undesirable outputs are compared, the source of inefficiency in each decision making unit (DMU) is calculated using scale efficiency, and the evolution over time is analyzed using the Malmquist index. The results indicate that the industries located in the Chilean regions of Coquimbo, La Araucania and Aysen were the most efficient while the industries in the regions of Tarapaca, Antofagasta and Biobio were less efficient. The most efficient industrial sectors were those involving communications equipment, base metals, and clothing; the least efficient were those concerned with food and beverages, textiles and nonmetallic minerals. Due to the treatment of the undesirable outputs, differences were found in the efficiency indexes obtained by the three models. This finding suggests using a model better adapted to the characteristics of the outputs in question and the viability of improving industrial practices. •An environmental efficiency analysis of Chilean manufacturing sectors is developed.•DEA with Malmquist index are used for the environmental assessment in a time horizon.•Three approaches to deal with undesirable outputs in DEA are compared.•The most efficient sectors were communications equipment, base metals, and clothing.•Differences were found in the efficiency indexes obtained by the three approaches.