This article deals with the influence of working memory features on translation performance and reports on one of the studies carried out as part of a PhD research designed to investigate the relationship between WM and performance in translation, copying simple clauses in L1, writing in L1, and reading in L2. The research points out that regression models used to verify causal relations did not show a significant influence of WM on translation. However, experts and novices differed as to how their results related to each other. Evidence showed that task coordination among the subjects related to typing, whereas processing speed related to measures tapping into higher processes during translation. No evidence was found of the importance of storage capacity. It is reasonable to suppose that translation experience does not imply acquiring a completely new ability, but rather organizing a more efficient, and resource-saving way of approaching the translation task.