Long length homogeneous meteorological observations are required for various hydro-meteorological applications, develop forecasting models for mitigation of hydro-meteorological hazards, climatic change and its impact studies, planning for sustainable developments and climate change adaption etc. Such observations for the high-altitude mountainous areas such as the North-West Himalaya (NWH) (or elsewhere) are not available due to the inhospitable climatic conditions and complex topography. The reanalysis (or refined analysis) data provide estimates of the meteorological variables. However, it has been reported that the reanalysis (or refined analysis) data exhibit errors and biases for meteorological variable(s). The errors and biases in the reanalysis (refined) analysis meteorological variable(s) can be reduced and long length data can be generated efficiently via statistical downscaling. In this study, three statistical downscaling methods; altitude correction (ALTC), regression (SR), quantile–quantile mapping (SQ), are developed and employed to statistically downscale daily mean air temperature of the High Asia Refined Analysis version 2.0 (HARv2) at 10 stations in the NWH, India and their performances are evaluated and compared. The daily mean air temperature of the HARv2 is found to exhibit statistically significant positive correlation (CC) with the observed daily mean air temperature (OB) at each station. This suggests suitability of the ALTC, SR and SQ for statistical downscaling of daily mean air temperature of the HARv2.The root mean square error (RMS) for the estimation of the observed daily mean air temperature of the HARv2 data is found to fall in the range 5.7–11.6 (6.3–11.9) ℃ and it is found in the range 5.6–9.1(6.3–8.6) ℃, 4.2–7.8 (3.9–6.4) ℃, 4.2–7.8 (3.9–6.4) ℃, 4.2–7.8 (3.9–6.4) ℃, and 4.2–7.8 (3.9–6.4) ℃ for the training (test) data sets at 10 stations in the NWH. These results show that the statistical downscaling improves estimation of the observed daily mean air temperature in the NWH and it is fairly possible to develop long length data on daily mean air temperature utilizingHARv2 at a specific location in the NWH. Developed long length data on daily mean air temperature can be useful for many applications such as hydro-meteorological applications and forecasting, study of climatic change and its variability, disaster mitigation, planning for sustainable development in the NWH.