Recent work on the timing of proboscidean tooth enamel formation has led to advances in the use of stable isotopes to understand diet and behavior (Dirks et al 2012, Metcalfe et al 2011). Although these techniques continue to yield excellent insights into the diet and behavior of mastodons and mammoths (e.g., Metcalfe and Longstaffe 2012), sampling techniques typically require the removal of (sometimes large) sections of an enamel ridge-plates. We have modified this technique to accommodate in situ micro-sampling of complete specimens. Precise movement of the specimen in three dimensions is controlled by a Newmark NSC-G, 3-axis motion controller. A 4 cm diameter ball joint is mounted to this stage, and used to level the large metal plate (70 cm x 100 cm) that holds the specimen. The specimen is affixed to this plate with orthopedic thermoplastic. Thermoplastic softens above 70° C. Once heated, a cradle can be molded to conform to irregular specimen shapes and mounted to the metal plate using machine screws. A Proxxon 50/E drill equipped with a 500 um diameter burr bit is affixed to a stationary arm above the 3-axis stage. Each specimen is sampled in multiple sets which each correspond to 1 cm of tooth growth. Each 1 cm set consists of ten samples and each sample consists of a series of vertical sub-samples. Each 100 um deep sub-sample is individually collected and processed through the entire thickness of the enamel. All enamel powder is collected with a micro-pipetter in de-ionized water to 1) maximize sample recovery, and 2) lubricate the bit. The sample nearest the enamel-dentin junction is analyzed for δ13C, δ18O or 87Sr/86Sr for paleoecological information. Although this technique is both time- and labor-intensive, it is minimally invasive and is capable of sampling enamel growth structures at a high resolution.