The present study was aimed at developing and assessing an automatic prompting instrument with three multihandicapped students who showed breaks in performance. The effects of the instrument were compared with the effects of manual prompting with simple tasks involving locomotion. The results indicated that manual and instrument prompting were both successful in reducing the amount of time spent in breaks. The effects of the instrument were replicated within subjects. The same effects were still visible during a follow-up assessment. The reliability and durability of those effects as well as the practical implications of automatic prompting are discussed.