This dissertation provides a continuous, parametric study of the development of task-switch efficiency between the ages of 3-7 years. It examines the degree to which several leading explanations for switching efficiency from both the developmental and adult literature predict individual and age-related gains in accuracy and reaction time (RT) in switching tasks in early childhood. These include executive functioning (executive functions–based) accounts along with newer models of cue-processing effects. Previous work has recognized the importance of cue-processing in adult task-switch, but this dissertation provides one of the first systematic investigations of cue-processing effects in childhood. We manipulate the semantic content of verbal cues both in task-switch paradigms and in simple, no-conflict, non-switch cued matching tests. We find that differences integrating such cues can account for both individual differences and age-related changes in flexibility. These include the catastrophic switching errors made by young children in traditional task-switch tests (and also made by older children in more complicated, indirectly or arbitrarily-cued paradigms) as well as children’s switch costs. Traditional explanations for performance on card-based rule sorting tasks (such as the Dimension Change Card Sort Task, DCCS (Zelazo, 1996, 2006) do not adequately account for both performance patterns. Further, we extend the idea of cue-processing and cognitive flexibility as predictors of one kind of real-world indirect cue: discourse referents.