Trace element analysis of high Z accessory phases by electron probe micro-analysis offers unique access to data unattainable by other analytical techniques, but also presents great challenges. The intimate relationships between beam conditions (voltage, current), spatial resolution, precision, and spectral complexity require the careful examination of all analytical parameters and their consequences, at a level commensurate with the sensitivity of the desired analysis. The high REE and actinide content, and resulting high Z of monazite, allows high direct analytical resolution (below 500 nm at 10 kV, depending on the realized beam diameter) due to limited electron range, but the high concentration of these elements will also produce significant fluorescence at a distance. The use of high accelerating potential in an effort to improve counting statistics is injudicious in most cases, as direct analytical resolution decreases, and high overvoltage on REEs and Th in monazite results in efficient production of energetic ionizing radiation (L-series) capable of fluorescing elements of interest, or interfering elements at remarkable distances (at least several 10 s of microns from the beam). Boundary fluorescence is a very significant problem in trace element analysis and in geochronologic analysis of monazite, as grains are commonly complex, with micro-domains that differ considerably in composition (especially Th/REE). In such instances, a 15 kV beam can produce fluorescence at a distance, leading to errors of 10 s of ppm within 5 μm of a compositional boundary. For trace element analysis, effects such as background curvature and minor interferences can result in very large errors, 50% or more for curvature alone for concentrations at or below 100 ppm. Strong background curvature is observed for all makes and varieties of spectrometers in the Pb-M region. In addition, for REE and actinide bearing phases such as monazite, the wavelength regions available for background determination are very limited, requiring routine, high-precision scanning and regression modeling to determine reasonably accurate background intensities. Two-point background interpolation methods are inappropriate for high sensitivity analysis in all cases. Without a direct means to verify the accuracy of low concentration analyses, great attention must be paid to analytical protocol and very detailed spectral analysis. The use of consistency standards to monitor changes is important, but caution must be exercised in the use of secondary “age” standards in the case of monazite EPMA because systematic error can result in cancellation effects when concentrations of both parent and daughter are proportionally incorrect. In all cases, caution in interpretation and integration of all available information is necessary if the potential of the method is to be realized.