Recent publications have reported increased signal intensities (SI) in the brain regions dentate nucleus (1-7) and globus pallidus (1,3,5) on unenhanced T1-weighted images after repeated administrations of gadolinium based contrast agents (GBCAs). The aim of the present animal study was to reproduce the reported clinical observations in a rat model after repeated, high-dose administrations of linear and macrocyclic GBCAs using analogous imaging and evaluation methods.Sixty healthy Han-Wistar rats were divided into a control and five GBCA groups (10 animals per group). The animals received either one linear GBCA (gadodiamide, gadopentetate dimeglumine, or gadobenate dimeglumine) or one macrocyclic GBCA (gadobutrol or gadoterate meglumine). Over a time period of two weeks the animals received 10 intravenous injections each at a dose of 2.5 mmol Gd/kg body weight on five consecutive days per week. The control group received identical volumes of saline. The GBCAs were administered by slow hand injection in their marketed formulation. Before GBCA administration, as well as three and twenty-four days after the last injection, a whole brain MRI was performed using a T1-weighted 3D Turbo-Spinecho sequence (TR/TE = 500/19 ms) on a clinical 1.5T scanner equipped with a dedicated rat head coil. The ratios of signal intensities in cerebellar nuclei to pons (CN/Po) and globus pallidus to thalamus were analyzed by two experienced readers.The experimental groups that received the linear GBCAs gadodiamide, gadobenate dimeglumine and gadopentetate dimeglumine showed enhanced SI of the cerebellar nuclei compared to baseline (Fig.1). The evaluation of the CN/Po SI ratio exhibited a statistically significant increase for gadodiamide and gadobenate dimeglumine three and twenty-four days after the last injection (Fig.2). No significant changes were observed between the two time points. Gadopentetate dimeglumine injection led to a moderate, but statistically not significant, increase of the CN/Po ratio. In the MRI images of rats that received the macrocyclic GBCAs gadobutrol and gadoterate meglumine or saline no increased CN/Po SI ratios were observed. The ratio of SI in globus pallidus versus thalamus was not elevated in any rats injected with GBCAs or saline (Fig.2).Differences exist between linear and macrocyclic GBCAs with respect to increased SI in the cerebellar nuclei in this animal model. Enhanced SI of the cerebellar nuclei were found after multiple, high-dose administrations of linear GBCAs. In contrast, no changes of SI were found for the investigated macrocyclic GBCAs. The clinical observation of increased SI in the globus pallidus after administration of linear GBCAs was not reproduced in this animal model.