It has recently been recognized that accessory minerals in continental rocks may contain a substantial fraction of the whole-rock complement of trace elements and geochemically important isotopes. This observation is in itself interesting, but its importance is magnified many fold when one considers the possible consequences to the geochemical rearrangement that accompanies crustal melting. If, for example, an accessory phase is insoluble in early-formed partial melts, it may effectively retain in the residue many incompatible elements (e.g., REE) and daughter isotopes (e.g., radiogenic Pb) that would otherwise be partitioned into the melt. In addition, should an unmelted accessory mineral (e.g., zircon) be entrained in a departing melt fraction, it may carry with it radiogenic components (Pb∗, Nd 143, Hf 176) characteristic of the source region. The effectiveness of phases such as zircon, apatite, sphene, allanite-epidote, and monazite in the roles noted above can only be gauged from knowledge of the three “fundamental accessory-phase parameters,” i.e. (1) the solubilities of the accessory minerals in crustal melts; (2) the equilibrium mineral/liquid partition coefficients for the trace element and isotopes of interest; and (3) the diffusivities that govern the rates at which equilibrium will be approached. In general, these parameters can only be measured in laboratory experiments, and relatively few data have been accumulated to date. It does seem clear, however, that apatite generally is residual (insoluble) during crustal melting and zircon can be, depending on melting conditions and abundance of zircon in the source. Petrographic and geochemical studies indicate that monazite, allanite, and epidote may also be highly insoluble in some cases, but the specific circumstances that promote insolubility remain unknown. The few experimentally-determined partition coefficients now available for apatite and zircon agree qualitatively with natural rock data on mineral separates. However, preliminary diffusion measurements and inferences from natural systems suggest that trace-element and isotopic equilibrium between undissolved accessories and crustal melts is in fact rarely approached. Ironically, it is this very disequilibrium in the form of preserved diffusion profiles, that may afford the most information on time-temperature histories of melting events. Continued experimental interest in the fundamental accessory-phase parameters will eventually provide the tools that will enable rigorous assessment not only of the timing and duration of crustal melting episodes, but also of the extent of melting and re-distribution of trace elements and radiogenic isotopes.