RATIONALE AND OBJECTIVES: In vivo tracking cells using gadolinium-based contrast agents have the important advantage of providing a positive contrast on T1-weighted images, which is less likely to be confused with artifacts because of postoperative local signal voids such as metal, hemorrhage, or air. The aim of this study is to paramagnetically and fluorescently label marrow with dual agents (gadolinium-diethylene triamine penta-acetic acid [Gd-DTPA] and PEI-FluoR) and track them after transplantation into spinal cord injury (SCI) with magnetic resonance imaging (MRI). MATERIALS AND METHODS: Marrow mesenchymal stem cells (MSCs) from Sprague-Dawley rats were incubated with PEI-FluoR (rhodamine-conjugated PEI-FluoR) and Gd-DTPA complex for labeling. After labeling, cellular viability, proliferation, and apoptosis were evaluated. T1 value and longevity of intracellular Gd-DTPA retention were measured on a 1.5 T MRI scanner. Thirty-six SCI rats were implanted with labeled and unlabeled MSCs and phosphate-buffered saline. Then, serial MRI and Basso-Beattie-Bresnehan (BBB) locomotor tests were performed and correlated with fluorescent microscopy. The relative signal intensity (RSL) of the engraftment in relation to normal cord was measured and the linear mixed model followed by post-hoc Bonferroni test was used to identify significant differences in RSL as well as BBB score. RESULTS: MSCs could be paramagnetically and fluorescently labeled by the dual agents. The labeling did not influence the cellular viability, proliferation, and apoptosis. The longevity of Gd-DTPA retention in labeled MSCs was up to 21 days. The distribution and migration of labeled MSCs in SCI lesions could be tracked until 7 days after implantation on MRI. The relative signal intensities of SCI rats treated with labeled cells at 1 day and 3 days (1.34 +/- 0.02, 1.27 +/- 0.03) were significantly higher than rats treated with unlabeled cells (0.94 +/- 0.01, 0.99 +/- 0.02) and phosphate-buffered saline (0.91 +/- 0.01, 0.95 +/- 0.01) (P < .05). Rats treated with labeled MSCs or unlabeled MSCs achieved significantly higher BBB scores than controls at 14, 21, 28, and 35 days after injury (P < .05). CONCLUSIONS: Labeling MSCs with the dual agents may enable cellular MRI and tracking in experimental spinal cord injury.