Interpolation of elevation data is a common step for creation and reformatting of many digital elevation products, which involves the estimation of unknown surface values based on surrounding measurements. Hydrological and landscape modelling applied using geographical information systems (GIS) commonly involves use of interpolated DEMs, in which inaccuracies in elevation values may lead to propagation of larger errors in derived terrain values. Because of this, assessment of errors or inaccuracies in these products is vital to ensure that the data quality is sufficient for the given purpose. This study firstly explores the quality of DEMs interpolated and resampled using a number of common techniques, and explores the effects that varying terrain morphology and sampling density have on the outcome of the generated products. This research shows how terrain morphology and sample density are first order controls on the accuracy of interpolated DEMs, and demonstrate that by quantifying terrain ruggedness using a common global index, the RMSE of elevation and RMSE of terrain derivatives at a global scale can often be fairly estimated. Secondly, the propagation of elevation error into terrain calculations is also explored, showing how RMSE of elevation can be an excellent predictor of global error in derived terrain values.