In contrast to the depth of knowledge available for the enhancement of plant species diversity and ecosystem services through ecological restoration, our understanding of how ecological restoration impacts genetic diversity (GD) of plant species has not yet been synthesized. We performed a global meta‐analysis to examine whether ecological restoration improved GD of plant species in restored populations. First, we compared the GD of restored populations with reference or degraded populations. Second, we explored whether the influence of ecological restoration on plant GD varies between species with different characteristics (life form and threat status), between different restoration strategies (active/passive, seeding/planting, mixture/non‐mixture) or between different restoration times (<50 and ≥ 50 years; with an average of 29.3 years). The GD of restored populations was significantly lower (HE, 1.06%; PPB, 5.10%, and SWI, 4.95%) than in reference populations but was comparable to degraded populations. The inbreeding coefficient (FIS, the proportion by which the heterozygosity of an individual is reduced by inbreeding) was consistently comparable between restored populations and reference or degraded populations. Woody species but not herbs and forest but not grassland ecosystem had significantly lower GD in restored populations than in reference populations. Passive but not active restoration, seeding rather than planting, and mixing materials from different sources rather than using a single source, all significantly increased the GD of restored populations. When the restoration time was ≥50 years, in contrast to <50 years, GD was comparable between the restored and reference populations. Synthesis and applications. In general, ecological restoration did not significantly improve the GD of plant species compared to reference or degraded populations. This might be due in part to the relatively short restoration time. Using passive restoration, seeding, and mixed sources could significantly increase the GD of restored populations. We emphasize that GD should not be treated as a minor cobenefit of ecological restoration for other purposes and that the recovery of GD should be listed as a vital goal in future ecological restoration with plant species. 摘要 虽然生态恢复对植物物种多样性和生态系统服务的促进作用已得到深入的研究和综述，但是生态恢复对植物遗传多样性的恢复效果尚未得到系统的总结。 本研究采用Meta分析来揭示全球范围内生态恢复是否促进被恢复植物种群的遗传多样性。首先，我们比较了被恢复植物种群与参照或退化种群的遗传多样性。其次，我们探索了植物特性(生活型和受威胁状况)、恢复策略(主动与被动修复、直接播种与活体种植、多种源与单种源)和恢复时间(小于与不小于50年；平均恢复时间为29.3年)对植物遗传多样性恢复效果的影响。 被恢复植物种群的遗传多样性显著低于参照种群(预期杂合度HE，低1.06%；多态性位点百分比PPB，5.10%；香农‐维纳指数SWI，4.95%)，但是与退化种群相当。被恢复植物种群的近交系数(FIS)与参照和退化种群均相当。 木本而非草本植物、森林而非草地生态系统的被恢复植物种群的遗传多样性显著低于参照种群。被动而非主动恢复、直接播种而非活体种植、多种源混合而非单种源可以显著提高被恢复植物种群的遗传多样性。当恢复时间小于50年时，被恢复植物种群的遗传多样性显著低于参照种群，而当恢复时间大于等于50年时，被恢复种群的遗传多样性与参照种群相当。 总的来说，与参照和退化种群相比，生态恢复并未显著提高被恢复植物种群的遗传多样性，部分原因可能是生态恢复时间相对较短。在条件允许的情况下，采用被动恢复、直接播种和多种源混合的策略可以显著提高被恢复植物种群的遗传多样性。本研究强调，在未来涉及植物的生态恢复中，需将植物种群的遗传多样性恢复作为一个重要目标。 In general, ecological restoration did not significantly improve the GD of plant species compared to reference or degraded populations. This might be due in part to the relatively short restoration time. Using passive restoration, seeding, and mixed sources could significantly increase the GD of restored populations. We emphasize that GD should not be treated as a minor cobenefit of ecological restoration for other purposes and that the recovery of GD should be listed as a vital goal in future ecological restoration with plant species.