Objective: A new imaging modality, X-ray Luminescence Tomography (XLT), is being developed for molecular imaging. This modality uses a collimated X-ray source to image the concentration of nano-sized phosphors tagged with molecular targets. This technology utilizes a limited-angle raster-scanned source and a projection-type detection geometry which is amenable to surgical interventions. This study compared the resolution and contrast recovery of XLT and fluorescence molecular tomography (FMT). Methods: Simulations and phantom measurements were performed for comparison. An X-ray model was used for excitation modeling (for the XLT case), and a diffuse optical light propagation model for emission (and reflection for the FMT case). Various inclusions with 5x contrast were imaged in a 9 x 5 cm volume. The inclusions were placed at various depths to determine the concentration recovery performance between XLT and FMT. Experimental measurements used a 50 keV X-ray radiotherapy source collimated into a 1mm-wide beam, and an EM-CCD camera. Tissue-simulating solutions of India ink and intralipid solutions were mixed in a 12 x 7cm phantom. Two cylindrical inclusions, 3mm and 9mm, of gadolinium oxysulfide: europium phosphors (625, 705 nm emission), sized between 10 and 25nm, were imaged at various depths. Results: For all inclusion sizes tested (2mm to 16mm diameters), recovered inclusion contrast decayed rapidly with depth for the FMT case; recovered contrasts were less than 10% accurate for depths greater than 15mm. XLT was able to distinguish the inclusion with at least 50% accuracy for depths up to 35mm for inclusion sizes greater than 6mm, and at least 25% accuracy for sizes down to 3mm. Image reconstructions from the experimental phantom were able to recover 50% contrast for the 3mm inclusion at depths up to 2.5cm, and the 9mm inclusion was at least 50% accurate at these depths. Discussion: X-ray excitation is highly localized due to lack of scatter. This is especially true when compared to FMT, where light is an diffuse field emitting from an isotropic source. Thus, spatial localization at depth is poor in FMT. Similarly, concentration recovery suffers at shallower depths. This could limit its use during surgical procedures, such as lumpectomy, where a spiculated lesion could extend well beyond the surgical margin, or if poorly resolved, could be masked by background heterogeneity. Future work will involve investigation in small-animals.