Landscape-scale ecological modelling has been hindered by suitable high-resolution surface meteorological datasets. To overcome these limitations, desirable spatial attributes of gridded climate data are combined with desirable temporal attributes of regional-scale reanalysis and daily gauge-based precipitation to derive a spatially and temporally complete, high-resolution (4-km) gridded dataset of surface meteorological variables required in ecological modelling for the contiguous United States from 1979 to 2010. Validation of the resulting gridded surface meteorological data, using an extensive network of automated weather stations across the western United States, showed skill comparable to that derived from interpolation using station observations, suggesting it can serve as suitable surrogate for landscape-scale ecological modelling across vast unmonitored areas of the United States.