The chapter discusses the question to what extent the effect of task complexity on linguistic performance in L2 writing and speaking is influenced by the mode in which the tasks have to be performed (oral versus written). The majority of studies which have been conducted so far in the framework of either the Limited Attentional Capacity Model (Skehan, 1998; Skehan & Foster, 2001) or the Multiple Attentional Resources Model (Robinson, 1995, 2007, this volume) concern oral task performance. While some studies have also looked at the effect of task complexity on the written production of L2 learners, there are no studies to our knowledge in which the effect of task complexity on linguistic performance in relation to mode has been investigated. For that reason a study was set up where two tasks of different task complexity, which already had been submitted to L2 learners in the writing mode, were presented to a group of L2 learners as speaking tasks. The participants in the oral mode were 44 learners of Italian as a second language, with Dutch as their mother tongue. Their performance was compared with that of another group of 91 Italian L2 learners with Dutch L1 who had performed the same tasks in the written mode. Scores on a cloze test were used as a measure of the general level of L2 proficiency of the learners. Our results demonstrate that both in the oral and the written mode task complexity mainly seems to affect accuracy, in particular with respect to lexical errors. We did not observe an interaction of task type and proficiency level, either in the written or in the oral mode.