Several techniques have been used to sample annual dentin layers in mammalian teeth for stable isotope analysis. However tooth size in smaller animals and the conical arrangement of the dentin layers have constrained precision of subsampling and collection of adequate sample from each annual growth layer. We tested an alternative subsampling technique using teeth from Risso's dolphins (n = 15) that involved cutting out the annual dentin growth layer groups (GLGs) from 300 to 500 μm longitudinal sections from one half of a demineralized tooth and comparing the results to those obtained using a standard micromilling process on the other half of the same tooth. Subsamples were analyzed for elemental C and N content and for stable C and N isotopes. Subsamples obtained from cutting out the GLGs showed more consistent wt%N, wt%C, and atomic C/N ratios that were significantly different (P<0.0001) from those obtained by micromilling. Consequently, the δ15N and δ13C values differed significantly between methods with values from the former method being more concordant with the expected variations in the early years of Risso's dolphins. Deviations in δ13C values in the micromilled subsamples were large enough to create possible errors in interpretation of dietary sources. Cutting out the dentin layers reduced sample processing time by 90% and yielded ∼10% more collagen than micromilling. These results suggest that cutting out the annual dentin layers can produce greater yield of samples of better collagen quality with a much shorter processing time than the micromilling process and is, therefore, an effective method to subsample small mammalian teeth.