This article draws on relevance theory (Sperber and Wilson 1986/1995) and its application to translation (Gutt 2000) to investigate processing effort in translation in relation to two different types of encodings, namely conceptual and procedural encodings (Blakemore 2002, Wilson 2011). Building on the experimental paradigm of data triangulation in translation process research (Alves 2003; Jakobsen 2005), it analyses the translation processes of eight professional translators when performing a direct and an inverse translation task. The analysis focuses on the number and types of encodings found in micro/macro translation units (Alves and Vale 2009; 2011). Results suggest that processing effort in translation is greater in instances of procedural than conceptual encodings.