Two experiments tested the roles of hidden toys and motor history in the AB task with 10-month- old infants. In Experiment 1 ( N = 24), infants were tested in lid and toy versions of the task, each comprised of A and B trials. No toys were ever hidden in the lid condition. On all A trials, an experimenter directed infants' attention to one of two lids (the A lid) and allowed infants to reach following a 5-s delay. On B trials in the lid condition, the experimenter directed infants' attention to the other, B lid. On B trials in the toy condition, the experimenter directed infants' attention to a toy that was then hidden underneath the B lid. Following a 5-s delay, infants reached perseveratively to A—producing the AB error—in the lid condition (replicating Smith, McLin, Titzer, & Thelen, 1995), but not in the toy condition. In Experiment 2 ( N = 24), infants were tested in similar lid and toy versions of the task, except that on all A trials the experimenter directed attention to a toy that was then hidden underneath the A lid. Infants produced AB errors across lid and toy conditions. Contrary to Smith et al.'s (1995) claims, these findings indicate that infants distinguish hidden toys from lids alone in the AB task. The presence of hidden toys on A trials and on B trials, not simply infants' motor history, influence production of the AB error.