The neural tolerability of iodixanol has been assessed in studies in mice and dogs. The animals received up to 4 injections in the cisterna cerebellomedullaris while under light anesthesia. Iotrolan was included as a reference substance in 1 study. The observations comprised assessment of clinical behavior, cerebrospinal fluid analysis, hematology, clinical chemistry and/or macroscopic and microscopic examination at necropsy. In addition, the repeated-dose dog study, urinalysis and ophthalmoscopy were performed, electrocardiograms obtained, and respiratory rate, blood pressure and rectal temperature measured. Clinical signs and minor pathological changes caused by the injection procedures were seen in all studies in some animals treated with iodixanol as well as in control animals. Single (2.0 g I/kg) and repeated (0.960 g I/kg) intracisternal administration of iodixanol to mice caused no significant toxicological effects. Two dogs treated with a high dose of iodixanol (0.256 g I/kg; 0.8 ml/kg) had pathological changes (meningeal inflammation and/or necrosis) that were more severe than those observed in control dogs. Single and repeated intracisternal administration of 0.128 g I/kg (0.4 ml/kg) of iodixanol to dogs, however, caused no significant toxicological effects. Apart from the findings in the 2 dogs, the neurological and neuropathological changes elicited by iodixanol were similar to those induced by control or reference substances. The results from these intracisternal toxicity studies in mice and dogs indicate a significant margin of safety regarding the use of iodixanol in clinical intra-thecal indications.