Recent research has revealed that the substitution of handwriting practice for typing may hinder the initial steps of reading development. Two hypotheses for the detrimental effect of typing are (a) reduced graphomotor activity and (b) reduced variability in the visual letter forms. However, previous studies were mostly limited to letter learning and primarily employed the visual identification of letters as a learning index. The current experiment investigated the impact of graphomotor action and output variability in letter and word learning using a variety of tasks. A total of 50 prereaders learned nine letters and 16 pseudowords made up of these letters across four learning conditions: copying the letters/words by hand, tracing the letters/words, typing the letters/words on a computer with several fonts, and typing with a single font. Posttest tasks included naming, writing, and visual identification of the trained letters and words. Results showed that children in the handwriting groups (i.e., trained through hand-copying or tracing) achieved higher accuracy across all posttest tasks compared with those in the typing groups. These outcomes illustrate the importance of handwriting experience in learning alphabetic and orthographic representations, favoring the graphomotor hypothesis. Thus, educators should be cautious about replacing pencil and paper with digital devices during the period of children's reading acquisition.