Rats were raised on a variety of isotopically controlled diets comprising 20% C 3, C 4 or marine protein and C 3 and/or C 4 non-protein or energy (i.e. sucrose, starch and oil) macronutrients. Compound specific stable carbon isotope (δ 13C) analysis was performed on the cholesterol isolated from the diet ( n=7 ) and bone ( n=15 ) of these animals and the values compared with bulk δ 13C measurements of bone collagen and apatite. The dietary signals reflected by these three bone biochemical components were investigated using linear regression analysis. δ 13C values of bone cholesterol were shown to reflect whole diet δ 13C values, collagen to reflect mainly dietary protein values and apatite to reflect whole diet values. Further correlations between dietary protein-to-energy spacings (Δ 13C prot-engy = δ 13C protein – δ 13C energy) and whole diet-to-bone component fractionations (Δ 13C bcomp- wdiet = δ 13C bone component – δ 13C whole diet) indicates that for hypothetical diets where protein δ 13C values are equal to energy values, fractionations between whole diet and bone biochemical fractions are –3.3‰ for cholesterol, +5.4‰ for collagen and +9.5‰ for apatite. Moreover, the narrow range of variation observed in apatite-to-cholesterol spacings (Δ 13C apat- bchol ) suggests that cholesterol δ 13C values can potentially also be used as an independent test for the isotopic integrity of apatite δ 13C values. These insights into bone cholesterol, collagen and apatite dietary signals, diet-to-bone fractionations and bone component-to-bone component spacings provide the basis for more accurate interpretations of the dietary behaviour of archaeological populations and food webs when the δ 13C analysis of bone is employed.