<p>The accuracy of hydrological assessments in mountain regions is often hindered by the low density of gauges, coupled with complex spatial variations in climate. Increasingly, spatial data sets (i.e. satellite and gridded products) and new computational tools are used to address this problem, by assisting with the spatial interpolation of ground observations. This paper presents a comparison of approaches of different complexity to spatially interpolate precipitation and temperature time-series in the upper Aconcagua catchment in central Chile. A Generalised Linear Mixed Model whose parameters are estimated through approximate Bayesian inference is compared with three simpler alternatives: Inverse Distance Weighting, Lapse Rates and a method based on WorldClim data. The assessment is based on a leave-one-out cross validation, with the Root Mean Squared Error being the primary performance criterion for both climate variables, while Probability of Detection and False Alarm Ratio are also used for precipitation. Results show that for spatial interpolation of the expected values of temperature and precipitation, the WorldClim approach may be recommended as being the more accurate, easy to apply and relatively more robust to tested reductions in the number of estimation gauges, particularly for temperature. The Generalised Linear Mixed Model has comparable performance when all gauges were included, but is more sensitive to the reduction in the number of gauges used for estimation, which is a constraint in sparsely monitored catchments.</p>