Abstract Portable X-ray fluorescence (pXRF) instruments can source obsidian artifacts once beyond analytical reach, expanding the range of artifact classes included in sourcing research. It is now more straightforward to analyze a “museum quality” tool on display in a gallery than the flaking chips and resharpening debitage recovered with it. That is because a minimum specimen size for artifacts is often mentioned as a drawback of pXRF-based sourcing. Many recent studies reject artifacts smaller than 1 cm and thinner than 3 mm for pXRF analysis. This study concerns whether such guidelines are too restrictive. After examining data from previous research into this issue, this study uses an experimental assemblage of 650 flakes from five obsidian sources to demonstrate two data handling techniques that enable valid and reliable source identifications for artifacts as small as microdebitage. The first involves ratios between quantitative concentrations of well-measured elements with similar X-ray energies, while the second one involves multivariate analysis of elemental data for obsidian flakes across a variety of size classes. As an example, 200 small blades are matched to two chemically similar obsidian sources on the Aegean island of Melos. Distinguishing between them has been treated as a “benchmark” or “litmus test” for evaluating obsidian sourcing techniques. Ultimately, it is demonstrated that debitage, as small as ∼ 1–2 mm and