Objective: Radioluminescent nanophosphors (RLNPs) show great promise for imaging biological processes in vivo under X-ray guidance, and for enhancing the efficacy and specificity of cytotoxic therapy during radiation therapy. As progress is made towards these goals, there is a need for an imaging modality that can quantitatively measure the distribution of RLNPs in small animals, with high sensitivity and high spatial resolution. X-ray luminescence computed tomography (XLCT) is best placed to meet all these requirements. In this scheme, collimated beams of X-ray radiation selectively excite RLNPs, producing near-infrared (NIR) light within a narrowly-defined volume. Optical measurement of the photons diffusing out of the subject can be interpreted as projective data and reconstructed into tomographic images. Methods: We have built a prototype XLCT system based on a first-generation CT geometry. The system is composed of a high-power KV X-ray source, a narrow tungsten collimator, a sensitive EM-CCD camera, and a computer-controlled motion stage. We have also synthesized NIR-emitting Eu3+-doped and green-emitting Tb3+-doped RLNPs, which were embedded in various phantoms and imaged with XLCT. A novel reconstruction scheme that includes a model of light propagation in biological tissue was developed and evaluated on XLCT scans with sparse angular sampling. Last, multiplexed imaging of RLNPs doped with different elements was demonstrated by resolving two different types of RLNP probes in a planar multispectral imaging set-up. Results: Imaging in an optically-diffusive medium shows that imaging performance is not affected by optical scatter; furthermore, the linear response of the reconstructed images suggests that XLCT is capable of quantitative imaging. Reconstruction combining models of light and X-ray propagation in tissue was found more accurate for sparse angular sampling. NIR- and green-emitting RLNPs were accurately resolved using two different imaging wavelengths. The RLNP distribution was estimated with as few as two projection angles. Conclusion: Based on phantom experiments, we found XLCT to be a feasible approach for imaging RLNPs tomographically. As we advance towards our goal of imaging RLNPs in live animals, we are designing and building a new imaging set-up with improved X-ray collimation and light collection efficiency, with simultaneous X-ray CT acquisition capabilities.