Despite nearly forty years of modern research on silicate melt inclusions (MI), only within the past 10-15 years have volcanologists and petrologists come to regularly accept their utility for characterizing magmatic systems. Their relatively slow acceptance was likely due to a number of factors including: 1) Lack of reliable analytical techniques, 2) Concern that MI represent anomalous boundary-layer melts or are altered by leakage or post-entrapment crystallization, 3) Data sets indicative of heterogeneous melts and, 4) Homogenization temperatures greater than those calculated by other techniques. With improvements in analytical methods and careful studies of MI systematics, workers are increasingly convinced of the utility of these features to unravel the complexities of volcanic systems: melt inclusions have "come of age." Recent studies provide compelling evidence for the compositions of dissolved and exsolved volatiles in magma reservoirs. Evidence for immiscibility of gases, hydrosaline brines and pegmatitic fluids demonstrate that magmatic phase relations are often more complicated than can be inferred by inspection of crystalline phases alone. ?? 2003 Elsevier B.V. All rights reserved.