From its inception, electron microprobe analysis was almost exclusively done under conditions of normal electron probe incidence. The radial symmetry of this geometry greatly simplified the development of quantitative equations, and these equations where further refined based on large amounts of data also collected at normal incidence. However, as x-ray detection systems where added to scanning electron microscopes (SEMs), samples were often viewed under conditions of non-normal incidence and attempts were made to modify the various correction procedures to give acceptable quantitative results. Little justification for these methods has ever been published and so the current study was undertaken to compare theoretically calculated x-ray emission from a well characterized sample, in this case NiAl (.685 wt. % Ni, .315 wt. %A1) with experimentally measured results collected as a function of tilt angle. The theoretical calculation where done using a Monte Carlo (MC) program developed by Gauvin and Lifshin.