Spatial climate data are often key drivers of computer models and statistical analyses, which form the basis for scientific conclusions, management decisions, and other important outcomes. The recent availability of very high-resolution climate data sets raises important questions about the tendency to equate resolution with realism. This paper discusses the relationship between scale and spatial climate-forcing factors, and provides background and advice on assessing the suitability of data sets. Spatial climate patterns are most affected by terrain and water bodies, primarily through the direct effects of elevation, terrain-induced climate transitions, cold air drainage and inversions, and coastal effects. The importance of these factors is generally lowest at scales of 100 km and greater, and becomes greatest at less than 10 km. Except in densely populated regions of developed countries, typical station spacing is on the order of 100 km. Regions without major terrain features and which are at least 100 km from climatically important coastlines can be handled adequately by most interpolation techniques. Situations characterized by significant terrain features, but with no climatically important coastlines, no rain shadows, and a well-mixed atmosphere can be reasonably handled by methods that explicitly account for elevation effects. Regions having significant terrain features, and also significant coastal effects, rain shadows, or cold air drainage and inversions are best handled by sophisticated systems that are configured and evaluated by experienced climatologists. There is no one satisfactory method for quantitatively estimating errors in spatial climate data sets, because the field that is being estimated is unknown between data points. Perhaps the best overall way to assess errors is to use a combination of approaches, involve data that are as independent from those used in the analysis as possible, and use common sense in the interpretation of results. Data set developers are encouraged to conduct expert reviews of their draft data sets, which is probably the single most effective way to improve data set quality. Copyright © 2006 Royal Meteorological Society.