Snow cover is one of the most important factors controlling microclimate and plant growing conditions for Arctic and alpine ecosystems. Climate change is altering snowfall regimes, which in turn influences snow cover and ultimately tundra plant communities. The interest in winter climate change and the number of experiments exploring the responses of alpine and Arctic ecosystems to changes in snow cover have been growing in recent years, but their outcomes are difficult to summarize because of the large variability in manipulation approaches, extents and measured response variables. In this review, we (1) compile the ecological publications on snow manipulation experiments, (2) classify the studies according to the climate scenarios they simulate and response variables they measure, (3) discuss the methods applied to manipulate snow cover, and (4) analyse and generalize the response in phenology, productivity and community composition by means of a meta-analysis. This meta-analysis shows that flowering phenology responded strongly to changes in the timing of snowmelt. The least responsive group of species were graminoids; however, they did show a decrease in productivity and abundance with experimentally increased snow covers. The species group with the greatest phenological response to snowmelt changes were the dwarf shrubs. Their abundance also increased in most long-term snow fence experiments, whereas species richness generally declined. We conclude that snow manipulation experiments can improve our understanding of recently observed ecosystem changes, and are an important component of climate change research.