Preclinical testing of new therapeutical strategies for the treatment of multiple myeloma (MM) requires animal models that closely resemble human disease and allow quantitative evaluation of the applied therapy. Models that meet both requirements have thus far not been described. Here we present a novel in vivo MM model by engraftment of MM U266 or RPMI-8226/S cells, both of human origin, into RAG2γc double knock-out mice. These mice are totally immune deficient because they lack T-, B and NK cells and the mice easily accept human cells (van Rijn et al., Blood 2003, Rozemuller et al., 2004). After intravenous injection of 2x106 MM cells engraftment and outgrowth occurred in all mice but was limited to the bone marrow compartment only. Flow cytometry (FCM) confirmed the presence of human CD45/38/138 positive MM cells in femur, spine, tibia and sternum bone specimens. Infiltration into other organs was not observed. In a next step MM cells were stably transduced using a retroviral vector encoding both the Green Fluorescent Protein (GFP) and firefly Luciferase (fLuc) marker genes. Technical advances in recent years in optical imaging by Bioluminescence Imaging (BLI) techniques allow visualization and quantification of bioluminescent light by detecting photons that are transmitted through mammalian tissue. When luciferase converts the substrate luciferin, photons are emitted that can be registered by using sensitive CCCD cameras. The absolute number of photons that are produced correlates, in our application, with local tumor mass. Mice were injected i.v. with 2x106 GFP-fLuc transduced MM cells (U266 or RPMI8226/S) and imaged weekly using BLI. Within 2 weeks after injection significant BLI signals were detectable. Per mouse 5-10 foci showed luciferase activity, predominantly in the pelvic region, skull, limbs, sternum, ribs and the spine. This low frequency of engraftment is in line with earlier reports on RPMI8226/S (Mitsiades et al., Cancer Res 2003). At 9 weeks the first mice developed hind leg paralysis which could be attributed to tumor associated spinal lesions. After 12 weeks the last mouse was sacrificed. BLI revealed that the intensity of light production at the various sites of tumor growth within individual mice as well as between mice showed a similar increase. This reflects an increase in tumor mass. Quantitative analysis of subsequent BLI images allowed construction of tumor growth curves of the total tumor mass per mouse as well as for the individual foci of MM growth in individual mice. We typically observed exponential growth, with growth curves running parallel with an average population doubling time of approximately 4–5 days. The range in which tumor growth could be monitored (and as a consequence also the response to treatment) spans 3–4 decades. In contrast with previously reported murine models for human MM where -next to bone marrow homing- also extra-skeletal tumors were observed our model almost exclusively shows homing of MM cells to the BM and is therefore more consistent with the clinical manifestation in myeloma patients. The major advantage of the model is the option for quantitative evaluation of the effect of a given treatment on the tumorload. Currently we are studying the efficacy of newly developed geranyl-geranyl-transferase inhibitors (GGTI).In conclusion, we have developed a novel in vivo model to study the characteristics of homing and outgrowth of MM and for quantitative evaluation of the efficacy of the therapeutic intervention applied.