Methods for measuring rates of DNA synthesis, and thus cell proliferation, in humans had not been available until recently. We (D. C. Macallan, C. A. Fullerton, R. A. Neese, K. Haddock, S. S. Park, and M. K. Hellerstein, 1998, Proc. Natl. Acad. Sci. USA 95, 708–713) recently developed a stable isotope–mass spectrometric technique for measuring DNA synthesis by labeling the deoxyribose (dR) moiety of purine deoxyribonucleotides through the de novo nucleotide synthesis pathway. The original analytic approach had limitations, however. Here, we describe technical improvements that increase yield, stability, sensitivity, and reproducibility of the method. The purine deoxyribonucleoside, deoxyadenosine (dA), is directly isolated from hydrolysates of DNA by using an LC18 SPE column. Two derivatives were developed for analyzing the dR moiety of dA alone (without the base), an aldonitrile-triacetate derivative, and a reduced pentose-tetraacetate (PTA) derivative. The PTA derivative in particular exhibited greater stability (no degradation after several weeks), greater GC/MS signal, and much less abundance sensitivity of isotope ratios (i.e., less dependence of mass isotopomer abundances on the amount of material injected into the mass spectrometer source), compared to previous derivatives of dA. The need for complex, multidimensional abundance corrected standard curves was thereby avoided. Using the PTA derivative, dR enrichments from DNA of fully turned over cells of rodents with ²H₂O enrichments in body water of 2.2–2.8% were 9.0–9.5%, and less than 1.0 μg DNA (ca. 2 × 10⁵ cells) was required for reproducible analyses. In summary, these methodologic advances allow measurement of stable isotope incorporation into DNA and calculation of cell proliferation and death rates in vivo in humans and experimental animals, with fewer cells, greater reproducibility, and less labor. Many applications of this approach can be envisioned.