In the present paper, the authors analyze a translation process implemented within the framework of Relevance Theory using Adaptation Theory as a tool to ensure relevance in the translation of popular science texts. The paper is part of ongoing research dedicated to the development of methodology for translation of popular science texts on architecture, ICT, and economics, focused here on translating from English into Latvian. Recognizing that relevance in translation is a qualitative category, the authors suggest measuring it along two dimensions: the plane of content and the plane of expression. Having defined four categories of relevance, the authors have developed a grid that may be recommended as a guide in a translator’s decision-making process for selecting a particular adaptive strategy and translation method.