Using co-speech iconic gestures (IG) supports L2 vocabulary learning in adults and children. The present study addressed this question in L1 with French kindergarten. Children had to learn 30 new words, 15 verbs and 15 nouns separated in 3 lists, through 3 conditions: using pictures (P) and using IG either by visualizing (V-IG) or by visualizing and reproducing IG (VR-IG). Words were presented through videos and training took place with groups of 3-4 children to reproduce the group dynamics of classroom. Participants were tested individually with word production (recall and denomination), word comprehension (recognition) and word meaning (definition) tasks immediately and 7 and 21 days after the end of the training. Preliminary results indicated that more words were significantly recalled and recognized in VR-IG than in P and V-IG conditions. This effect was also observed on the long-term measures. No beneficial effects of VR-IG over P and V-IG emerged for denomination task. Interestingly, for the recall task , we observed an interaction effect between conditions and vocabulary level of children assessed at the beginning of the study