It is challenging to understand the drivers of plant-plant interaction patterns in dry mountains. However, such knowledge is important to assess alpine ecosystem resilience to climate change. In water-limited ecosystems, leaf C-13 and satellite-derived vegetation index (NDVI) may serve as reliable indicators of environmental severity to address plant responses to water availability. We hypothesized that in dry mountains, interaction intensity between cushion and beneficiary species increases with increasing C-13 and decreasing NDVI regardless of elevation, indicating the importance of water availability in driving plant interactions. We used relative interaction indices (RII) of species and individual numbers within and outside the canopy of three cushion plant species along three elevational transects in dry Himalayas, Nepal. Site-specific NDVI was calculated from 30m Landsat images. Thornthwaite moisture index was calculated for each elevation site. We observed nonlinear patterns in RII, C-13, and NDVI with elevation. Intraspecific variation of C-13 was negatively correlated with moisture index and NDVI, while NDVI across sites was positively correlated with precipitation but not with temperature. RII within a cushion species was positively correlated with C-13 and negatively with NDVI when the effect of elevation was removed. In pooled data across cushion species and sites, RII was negatively correlated with precipitation and NDVI when the effect of temperature was removed. RII was uncorrelated with cushion size under the same environment. Leaf nitrogen showed no correlation with RII or C-13. Our data show that water shortage is the main driver of plant interactions in the alpine belt of dry Himalayas.