This study investigated alternate methods of letter discrimination pretraining and word recognition training in young children. Seventy kindergarten children were trained to recognize eight printed words in a vocabulary list by a mixed-list paired-associate method. Four of the stimulus words had visual response choices (pictures) and four had auditory response choices (spoken). Both methods involved a memory component; the stimulus and two response choices were presented sequentially. Method of instruction was also varied by two types of letter discrimination pretraining. Half of the students received discrimination pretraining in which response choices were presented visually and the other half received pretraining in which response choices were presented auditorily. The predictive measure of visual perception was the Frostig Developmental Test of Visual Perception, and the measure of auditory perception was a composite of the Memory subtest of the Test of Auditory Perception and the Sentences subtest of the Weschler Preschool Scale of Intelligence. The results indicated that three of the four predicted interactions between measures of perceptual ability and method of letter discrimination pretraining were not significant. One significant interaction was between visual perceptual ability and method of pretraining. (Author/WR)