This paper describes the work carried out within the scope of the TEMAA (Testbed study of Evaluation Methodologies: Authoring Aids) project and the EAGLES (Expert Advisory Groups on Language Engineering Studies) initiative. The aim was to lay the foundations for the standardisation of evaluation methodologies, applicable to a wide range of language engineering tools and evaluation scenarios. The theoretical background is based on the ISO standard for quality characteristics and guidelines for their use, concerning information technology and software product evaluation. This list of qualities is broken down into a set of attributes and values. The paper focuses on the notion of measurement involved in determining a value of an attribute and discusses validity and reliability. The evaluation of MT is looked at first from the perspective of internal and external validity (the latter being subdivided in time taken to post-edit and measuring comprehensibility); and, second, from the perspective of the end-user. The author’s personal views round off this paper.