Working memory, an essential component of cognitive function, can be improved through specific methods. This meta-analysis evaluates the effectiveness of transcranial alternating current stimulation (tACS), an emerging technique for enhancing working memory, and explores its efficacy, influencing factors, and underlying mechanisms. A PRISMA systematic search was conducted. Hedges's g was used to quantify effect sizes. We constructed a three-level meta-analytic model to account for all effect sizes and performed subgroup analyses to assess moderating factors. Recognizing the distinct neural underpinnings of various working memory processes, we separately assessed the effects on n-back tasks and traditional working memory tasks. A total of 39 studies with 405 effect sizes were included (170 from n-back tasks and 235 from other tasks). The overall analysis indicated a net benefit of g = 0.060 of tACS on working memory. Separate analyses showed that tACS had a small positive effect on n-back tasks (g = 0.102), but almost no effect on traditional working memory tasks (g = 0.045). Further analyses revealed mainly: A moderately positive effect of theta tACS (without anti-phase stimulation) on n-back tasks (g = 0.207); and a small effect of offline stimulation on working memory maintenance (g = 0.127). Overall, tACS has minimal impact on working memory improvement, but it shows potential under certain conditions. Specifically, both online and offline theta tACS can improve n-back task performance, while only offline stimulation enhances working memory maintenance. More research is needed to understand the mechanisms behind these effects to make tACS an effective method.