The visual short-term cued recall memory of 8-, 12-, and 16-month-old infants was assessed in 2 experiments using an operant indirect delayed-reaction procedure. Infants were first trained to touch a discriminative stimulus (a light) which appeared randomly in 1 of 2 locations for 3 sec. When infants reached a criterion performance of 5 consecutive correct responses, delay varying in duration were introduced between the offset of the light and the infant's touching response to the location at which it had appeared. In experiment 1, 8-month-olds and 12-month-olds were tested, using 250-msec and 3-, 6-, and 9-sec delays. Results indicated that the location of the discriminative stimulus could be recalled by 8-month-olds at 250-msec delays; and by 12-month-olds at 250-msec and 3-, 6-, and 9-sec delays. In experiment 2, 8-, 12-, and 16-month-olds performed better than chance, while 8-month-olds did not. The amount of visual attention was not related to the correctness of a response and was not used as a mediating strategy. Results are discussed with reference to previous research on the development of recall memory.