In this study, we report on the influence of trace elements (TE) on signal intensities of nuclear magnetic resonance images (MRI), both in vivo and in vitro. Optimal parameters for the assessment of Mn concentration in the brain of rats on total parenteral nutrition were established. For the in vitro study, Mn and trace element solutions, one containing Zn, Cu, Fe, and I (TE-4) and another containing the above elements plus Mn (TE-5), were diluted with physiological saline or with rat brain homogenate and used to measure signal intensities in MRI. Concentration-dependent signal hyperintensity was observed in both cases in the Mn and the TE-5 solutions, but no effect was observed with the TE-4 solution. The signal increase was greater for brain tissue homogenates. In the in vivo study, the experimental animals were maintained under total parenteral nutrition (TPN) with a standard clinical dose of TE-5 and/or with 10-fold the clinical dose of TE-4 and TE-5 for 1 wk. Only rats that were receiving the increased TE-5 dose showed signal hyperintensity on MRI. Positive correlations were observed among the signal hyperintensity, the blood Mn concentrations, and that of the rat brain. Our results suggest that Mn in TE preparations may be the cause of signal hyperintensity on MRI in a concentration-dependent fashion, and that MRI and measurement of blood Mn may be used to estimate Mn accumulation in brain tissue.