Residual switch cost refers to the RT and accuracy costs associated with switching tasks when there is ample opportunity to prepare for an upcoming switch (Rogers & Monsell, 1995). Although there is some debate, this component of switch cost likely represents a control process rather than a task execution process (Gopher, Armony, & Greenshpan, 2000). Recent evidence indicates that residual switch cost is greater for alternating tasks (e.g., ABA) than for non-alternating task switches (e.g., CBA), and provides one possibility for the control process underlying residual switch cost. Specifically, the alternating-task effect is hypothesized to reflect inhibition of a no-longer-relevant task-set, which would facilitate adoption of a currently-relevant task-set. When a participant must quickly reinstate the inhibited task-set, as when alternating between 2 tasks, additional cost is associated with resolution of the inhibition. The alternating-task effect is observed across variations of task type (i.e., perceptual & conceptual judgments), response modality (i.e., vocal & manual), and cue ambiguity. Individual differences such as working memory span and mild traumatic brain injury seem to have no effect on the magnitude of the alternating-task effect, but preliminary data indicates that individuals with ADHD may not show the effect, but instead show greater switch cost than matched controls. This pattern of results for ADHD indirectly supports the inhibitory hypothesis of the alternating-task effect, as ADHD is associated with compromised inhibition in other types of tasks (e.g., stop-signal tasks). The other manipulation that eliminates the alternating-task effect is combining task switches with changes of spatial location. This suggests that the task-set inhibition process may not operate when tasks can be uniquely associated with a spatial location, and that somewhat different processes may be involved in switching tasks across spatial locations.