Do second-language learners of Japanese organize mental lexicon referring to Japanese Kana script and process highly familiar Hiragana and Katakana words in the same way as Japanese native readers? In order to answer this question, an experiment employing a naming performance task was conducted. Foreign and native readers of Japanese were asked to read visually familiar and unfamiliar words written in Katakana and Hiragana. The results revealed a familiarity effect for loan words conventionally written in Katakana and for Japanese words written in Hiragana in foreign readers but not in native ones. This finding was explained in terms of differences in the structure of the mental lexicon, and in mechanisms involved in word recognition between the two groups of skilled and less skilled subjects. Namely, the results suggested that Japanese readers possess one internal lexicon referring to both Kana scripts, while foreign readers of Japanese make use of two separate lexicons in recognizing words written in Hiragana or Katakana.