This study compares second-year Japanese university students' strategies to write kanji by hand with their strategies to produce the kanji characters on a computer, taking into account factors such as accuracy in writing, the amount of kanji used, the complexity of the kanji used, as well as how the characters used compare with the sequence in which they are taught in the curriculum. Participants include 31 second-year second-semester college students, selected based on having studied two consecutive years of Japanese at the same university with the same textbook. Data sources include handwritten and typed dictations in Japanese, video screen captures of participants' computer use, handwritten and typed essay samples, interviews, and a questionnaire examining student attitudes towards the production of text in the typed and handwritten domains. Analysis of text production was based on Horodeck's (1987) kanji error analysis model with categorization further refined to account for the types of errors made by non-native speakers of Japanese. The results indicate that students tended to use more kanji with greater accuracy and tended to use complex kanji more frequently when typing in the dictation portion of the study. For the collected essays (232 handwritten and 232 typed essays), there was considerably less difference between the handwritten and typed output in terms of the kanji used and the accuracy thereof. Regarding the types of errors participants made in their production of kanji, typed production tended to produce homophone errors, due to the inherent system of typing in Japanese in which Romanized letters are used to produce the sounds of Japanese. Participants were frequently able to pick the right kanji combination even if they were not able to write that same combination by hand, although homophone errors in the typed domain were semantically more severe when they did appear. Participants were conservative in their use of kanji in handwritten tasks, while still producing orthographical errors, but did not produce phonological substitutions similar to those found in native speakers, suggesting that there is only a low level of phonological association with kanji at the participants' proficiency level. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]