Gravity reduction standards are needed to improve anomaly quality for interpretation and to facilitate the joining together of different data sets. To the extent possible, data reduction should be quantitative, objective, and comprehensive, leaving ambiguity only to the interpretation process that involves qualitative, subjective, and geological decisions. The term 'Bouguer anomaly' describes a field intended to be free of all nongeologic effects--not modified by a partial geologic interpretation. Measured vertical gradients of gravity demonstrate considerable variation but do not suggest, as often reported, that the normal free-air gradient is in error or needs to be locally adjusted. Such gradients are strongly influenced by terrain and, to a lesser extent, by the same geologic sources which produce Bouguer anomalies. A substantial body of existing literature facilitates the comprehensive treatment of terrain effects, which may be rigorously implemented with current computer technology. Although variations in topographic rock density are a major source of Bouguer anomalies, a constant density appropriate to the area under investigation is normally adopted as a data reduction standard, leaving a treatment of the density variations to the interpretation. A field example from British Columbia illustrates both the variations in vertical gravity gradients which can be encountered and the conclusion that the classical approach to data reduction is practically always suitable to account for the observed effects. Standard data reduction procedures do not (and should not) include reduction-to-datum. The interpreter must be aware, however, that otherwise 'smooth' regional Bouguer anomalies caused by regional sources do contain high-frequency components in areas of rugged topography.