This study compared effects of student choice of task sequence to two variations in teacher-manipulated task sequences on on-task behavior of elementary-aged students with disabilities. Researchers modified Call et al.'s (J Appl Behav Anal 42: 723-728, 2009) demand assessment to determine high-, moderate-, and low-probability tasks. Next, researchers applied the results from the demand assessment to inform teacher-manipulated variations in task sequences: a high- to low-probability task sequence and low- to high-probability task sequence. These sequences were then embedded in a visual activity schedule (VAS). Results of task sequence manipulation embedded in a VAS indicated slightly higher median percentages of on-task behavior for the high- to low-probability task sequence. Future directions for research based on these preliminary data are discussed.