The main purpose of the study was to investigate the effectiveness of the use of VNK and semantic processing on students’ word recognition and learning attitude. Eighty-four elementary sixth graders in New Taipei City participated in the study. They came from three classes, which served as three groups in this study, two experimental groups, Group A and Group B, and one control group. In group A, participants were trained to keep their vocabulary notebook in a semantically related way, and in group B, participants were required to keep their vocabulary notebook in an alphabetical order (semantically unrelated way). As for the control group, no training in notebook keeping or semantic processing was presented. The quantitative data were gathered from pretest, posttest, delayed posttest, and questionnaire, and analyzed with one-way ANOVA, two-way ANOVA and paired-samples t-test. As for the qualitative data received from interviews, it was discussed in terms of learners’ experience and preference. Although the results indicated that no significant differences existed among three groups on pretest, posttest and delayed posttest, Group B achieved significant improvement from pretest to posttest. In addition, participants’ opinions varied with their language proficiency level, especially in Group B. Most medium- and high-level students reported negative attitudes toward vocabulary notebook keeping, whereas all the low level learners showed great fondness. In contrast to Group B, all three proficiency levels in Group A showed almost the same preference for vocabulary notebook keeping.