We report the first high-precision characterization of molecular and intramolecular δ15N of nucleosides derived from mammalian DNA. The influence of dietary protein level on brain amino acids and deoxyribonucleosides was determined to investigate whether high protein turnover would alter amino acid 15 N or 13 C values. Pregnant guinea pig dams were fed control diets, or high or low levels of dietary protein throughout gestation, and all pups were fed control diets. The cerebellar DNA of offspring was extracted at 2 and 120 days of life, nucleosides isolated and δ15N and δ13C values characterized. Mean diet δ15N was 0.45 ± 0.33‰, compared with cerebellar whole tissue and DNA δ15N = +4.1 ± 0.7‰ and −4.5 ± 0.4‰, respectively. Cerebellar deoxythymidine (dT), deoxycytidine (dC), deoxyadenosine (dA), and deoxyguanosine (dG) δ15N were +1.4 ± 0.4, –2.1 ± 0.9, –7.2 ± 0.3, and −10.4 ± 0.5‰, respectively. There were no changes in amino acid or deoxyribonucleoside δ15N values due to dietary protein level. Using known metabolic relationships, we developed equations to calculate the intramolecular δ15N values originating from aspartate (asp) in purines (pur) or pyrimidines (pyr), glutamine (glu), and glycine (gly) to be δ15NASP-PUR, δ15NASP-PYR, δ15NGLN, and δ15NGLY +11.9 ± 2.3‰, +7.0 ± 2.0‰, –9.1 ± 2.4‰, and −31.8 ± 8.9‰, respectively. A subset of twelve amino acids from food and brain had mean δ15N values of 4.3 ± 3.2‰ and 13.8 ± 3.1‰, respectively, and δ15N values for gly and asp were 12.6 ± 2.2‰ and 15.2 ± 0.8‰, respectively. A separate isotope tracer study detected no significant turnover of cerebellar DNA in the first six months of life. The large negative δ15N difference between gly and cerebellar purine N at the gly (7) position implies either that there is a major isotope effect during DNA synthesis, or that in utero gly has a different isotope ratio during rapid growth and metabolism from that in adult life. Our data show that cerebellar nucleoside intramolecular δ15N values vary over more than 40‰ and are not influenced by dietary protein level or age. Copyright © 2011 John Wiley & Sons, Ltd.