In 2 experiments, the A-B, B-C and the A-B, C-A paradigms were compared with the A-B, C-D control using 2 levels of ability. For each experiment a different although equivalent sample of 120 Armor enlisted personnel was used (6 groups of 20 each). Both lists of Exp. I and List I of Exp. II utilized a modification of the anticipation procedure in which the stimuli were not pronounced. List II of Exp. II additionally used a multiple-choice procedure to equate response availability between groups. Original list learning unequivocally demonstrated the superiority of the high ability groups. List II comparisons showed clearly for the A-B, B-C paradigm that the high ability groups were faster but were somewhat equivocal for the A-B, C-A paradigm. The similar pattern of A-B, C-A transfer in both experiments was interpreted as suggesting the possibility of a source of facilitation operating on the perceptual level in Exp. II.