Inbound tourism has been paid much attention because it activates national and local economy in Japan, as the number of foreign tourists visiting Japan increased recently. This study investigates destination images of foreign tourists visiting Japan. This study examines with text mining technique two kinds of text, English travel guidebook (lonely planet) and tourist review site (tripadvisor). The text mining analyses included feature analysis and correspondence analysis. In conclusion, the authors found that the difference of destination images of foreign tourists from guidebook and review site was significant and that the images were also different according to the nationality of foreign tourist. The destination image should be modified if the difference might be far from the tourist expectation.