A recent extensive review of the ‘bioavailability’ of magnesium (Mg) (Henry and Benz 1995) noted that the fractional efficiency of apparent absorption of Mg (AAMg) in ruminants could vary from -0.04 to +0.66. Little attempt was made to identify sources of variation, other than those between inorganic Mg supplements. Earlier, ARC (1980) had failed to find consistent differences between fresh herbage (FH) and dry diets but a reexamination of the literature suggested that AAMg was twice as high for mixtures of roughage and concentrate (R/C) as for FH and confirmed the marked inhibitory effect of dietary potassium (K) (Suttle, 1987). Since grass is usually richer in K than a dry diet, differences between diet types in AAMg may be largely attributable to differences in K concentration. The objective of the study was to identify systematic sources of variation in the absorbability of dietary magnesium (Mg) to sheep from published balance studies.