High mountain ecosystems are vulnerable to the effects of climate warming and Australia’s alpine vegetation has been identified as particularly vulnerable. Between 2004 and 2010, we monitored vegetation changes in a warming experiment within alpine open grassy-heathland on the Bogong High Plains, Victoria, Australia. The study was part of the International Tundra Experiment (ITEX Network) and used open-topped chambers (OTC) to raise ambient growing-season temperatures by ~1°C at two sites. We assessed the effects of experimental warming on vegetation composition, diversity and cover using ordination, linear models and hierarchical partitioning. Results were compared with vegetation changes at four long-term (non-ITEX) monitoring sites in similar vegetation sampled from 1979 to 2010. The warming experiment coincided with the driest 13-year period (1996–2009) since the late 1880s. At the ITEX sites, between 2004 and 2010, graminoid cover decreased by 25%, whereas forb and shrub cover increased by 9% and 20%, respectively. Mean canopy height increased from 7 cm to 10 cm and diversity increased as a result of changes in relative abundance, rather than an influx of new species. These vegetation changes were similar to those at the four non-ITEX sites for the same period and well within the range of changes observed over the 31-year sampling period. Changes at the non-ITEX sites were correlated with a decrease in annual precipitation, increase in mean minimum temperatures during spring and increase in mean maximum temperature during autumn. Vegetation changes induced by the warming experiment were small rather than transformational and broadly similar to changes at the long-term monitoring sites. This suggests that Australian alpine vegetation has a degree of resilience to climate change in the short to medium term (20–30 years). In the long term (>30 years), drought may be as important a determinant of environmental change in alpine vegetation as rising temperatures. Long-term vegetation and climate data are invaluable in interpreting results from short-term (≤10 years) experiments.