In 1984, 29 obsidian artifacts and an unworked nodule from Tikal were attributed to source by visual means and then analyzed by X-ray fluorescence (XRF). We conclude that the considerable within-source optical variability of gray Mesoamerican obsidians makes visual sourcing unreliable at present, although a corpus of descriptions of the optical characteristics of obsidian may eventually provide a way to exclude possible sources. The XRF analysis identified two additional central Mexican sources, bringing to six the number of Mexican obsidian sources represented at Tikal; failed to provide a secure identification for the nodule; supported the succession of Guatemalan obsidian sources utilized by the Southern Lowland Maya previously proposed by Nelson; and suggested other correlations between source, artifact type, and archaeological context.