<p>There is a lack of clear consensus on the use of ‘substitution’ within attributional life cycle assessment (ALCA). A number of academic articles have previously argued that substitution is not consistent with ALCA, but a more recent article by Provost-Savard and Majeau-Bettez in the <i>Journal of Industrial Ecology</i> argues that it can be coherently used. In reply, this article contends that in order to resolve methodological disputes it is necessary to have a clear conceptualisation of what the method in question is intended to represent. To this end, the article clarifies that ALCA represents the processes used in the life cycle of the product that is studied, and also represents shares of total absolute anthropogenic impacts. A worked example for corn ethanol is used to show that substitution is fundamentally inconsistent with these features of ALCA, and to analyse what is right and wrong in Provost-Savard and Majeau-Bettez’s article. It is hoped that clarifying the reasons for why substitution is not appropriate in ALCA will help to build consensus on this issue. It is strongly recommended that standards and guidance for ALCA should clearly state that substitution should not be used.</p>