Three 60 m long, 2·8 m high snowfences have been erected to study long-term effects of changing winter snow conditions on arctic and alpine tundra. This paper describes the experimental design and short-term effects. Open-top fiberglass warming chambers are placed along the experimental snow gradients and in controls areas outside the fences; each warming plot is paired with an unwarmed plot. The purpose of the experiment is to examine short- and long-term changes to the integrated physical-biological systems under simultaneous changes of winter snow regime and summer temperature, as part of the Long-Term Ecological Research network and the International Tundra Experiment. The sites were at Niwot Ridge, Colorado, a temperate high altitude site in the Colorado Rockies, and Toolik Lake, Alaska, a high-latitude site. Initial results indicate that although experimental designs are essentially identical at the arctic and alpine sites, experimental effects are different. The drift at Niwot Ridge lasts much longer than do the Toolik Lake drifts, so that the Niwot Ridge fence affects both summer and winter conditions, whereas the Toolik Lake fence affects primarily winter conditions. The temperature experiment also differs in effect between the sites. Although the average temperature increase at the two sites is similar (daily increase 1·5 °C at Toolik and 1·9 °C at Niwot Ridge), at Toolik Lake there is only minor diurnal variation, whereas at Niwot Ridge the daytime increases are extreme on sunny days (as much as 7–10 °C), and minimum nighttime temperatures in the chambers are often slightly cooler than ambient (by about 1 °C). The experimental drifts resulted in wintertime increases in temperature and CO₂ flux. Temperatures under the deep drifts were much more consistent and warmer than in control areas, and at Niwot Ridge remained very close to 0 °C all winter. These increased temperatures were likely responsible for observed increases in system carbon loss. Initial changes to the aboveground biotic system included an increase in growth in response to both snow and warming, despite a reduced growing season. This is expected to be a transient response that will eventually be replaced by reduced growth. At least one species, Kobresia myosuroides, had almost completely died at Niwot Ridge three years after fence construction, whereas other species were increasing. We expect in both the short- and long-term to see the strongest effects of snow at the Niwot Ridge site, and stronger effects of temperature at Toolik Lake.