The purpose of this study was to determine effects of various prompting procedures in teaching children a sight vocabulary. The five words which comprised the sight vocabulary were presented to 96 five-year-olds in nursery and elementary schools. During initial training, integrated stimuli and non-integrated stimuli differed with respect to the spatial relation between the cue stimulus (CS) and eliciting stimulus (ES) in the CS+ES frame. In the prompt procedure, the learner's task was to select the correct word after seeing the ES frame and the ES+CS frame. In the confirmation procedure, the task was to select the correct word after seeing the ES frame but prior to seeing the ES+CS frame. During prompt overtraining (POT), the ES+CS frame was dropped, whereas in the -POT conditions, the ES+CS and the ES frames were dropped. Subjects were required to practice beyond the trial of last error for 10 additional trials. One week and eight weeks following overtraining to 10 criteria, the subject was given a retention test on the five printed names. Results for initial training suggest that prompts decreased the errors associated with learning the five new words. The data show that the integrated prompt technique decreased errors more than the non-integrated prompt technique. In terms of delayed retention, results suggest that errors are interfering only when they occur early in training. (Author/DB)