In this paper we describe a taxonomy of task demands which distinguishes between Task Complexity, Task Condition and Task Difficulty. We then describe three theoretical claims and predictions of the Cognition Hypothesis (Robinson 2001, 2003b, 2005a) concerning the effects of task complexity on: (a) language production; (b) interaction and uptake of information available in the input to tasks; and (c) individual differences-task interactions. Finally we summarize the findings of the empirical studies in this special issue which all address one or more of these predictions and point to some directions for continuing, future research into the effects of task complexity on learning and performance. (Contains 1 figure.)