The integration of physical movements, such as gestures, into learning holds potential for enhancing foreign language (L2) education. Uncovering whether actively performing gestures during L2 learning is more, or equally, effective compared to simply observing such movements is central to deepening our understanding of the efficacy of movement-based learning strategies. Here, we present a meta-analysis of seven studies containing 309 participants that compares the effects of gesture self-enactment and observation on L2 vocabulary learning. The results showed that gesture observation was just as effective for L2 learning as gesture enactment, based on free recall, cued L2 recognition, and cued native language recognition performance, with a large dispersion of true effect across studies. Gesture observation may be sufficient for inducing embodied L2 learning benefits, in support of theories positing shared mechanisms underlying enactment and observation. Future studies should examine the effects of gesture-based learning over longer time periods with larger sample sizes and more diverse word classes.