Objective chemical biomarkers are needed in clinical studies of diet-related diseases to supplement subjective self-reporting methods. We report on several critical experiments for the development of clinically legitimate dietary stable isotope biomarkers within human blood. Our examination of human blood revealed the following: (1) Within blood clot and serum from anonymous individuals (201 males, 205 females) we observed: mean serum δ13C = −19.1 ± 0.8‰ (standard deviation, SD); clot, −19.3 ± 0.8‰ (SD); range = −15.8‰ to −23.4‰. Highly statistically significant differences are observed between clot and serum, males and females for both clot and serum. For 15N (n = 206), mean serum = +8.8 ± 0.5‰ (SD); clot +7.4 ± 0.4‰ (SD); range = +6.3‰ to +10.5‰. Blood serum is enriched in 15N relative to blood clot by +1.4‰ on average, which may reflect differing protein amino acid content. Serum nitrogen is statistically significantly different for males and females, however, clot shows no statistical difference. (2) Relative to clot, capillary blood is marginally different for 13C, but not 15N. Clot 13C is not significantly different from serum; however, it is depleted in 15N by 1.5‰ relative to serum. (3) We assessed the effect of blood additives (sodium fluoride and polymerized acrylamide resin) and laboratory process (autoclaving, freeze drying) commonly used to preserve or prepare venous blood. On average, no alteration in δ13C or δ15N is detected compared with unadulterated blood from the same individual. (4) Storage of blood with and without the additives described above for a period of up to 115 days exhibits statistically significant differences for 13C and 15N for sodium fluoride. However, storage for unadulterated blood and blood preserved with polymerized acrylamide resin does not change the δ13C or δ15N isotopic composition of the blood in a significant way. With these experiments, we gain a clinical context for future development of a stable isotope based dietary biomarker. Copyright © 2008 John Wiley & Sons, Ltd.