Two groups of nine non-reading retarded children were trained with a paired-associates technique to compare the effects of a fading procedure and a standard (simultaneous presentation ) method on reading vocabulary acquisition. For the first 8-word list, the group received either faded or non-faded training and were tested to determine the number of words learned. Both groups were then given non-faded training on a second 8-word list and were again tested. The results indicated that the group which was given faded training on the first word list learned reliably more words from both lists than the non-faded-training group. Educational significance was seen in the fact that the fading procedure could itself be faded. The efficacy of fading was interpreted to be consistent with an attentional explanation : the salience of the picture stimuli was reduced by fading, which increased the relative salience of the words.