A rabbit model was used to compare the effect on the blood-brain barrier of the intracarotid injection of two new contrast media: iohexol, a nonionic monomer, and iodixanol, a nonionic dimer. It was hypothesized that the low osmolality of iodixanol (272 mOsm/kg at 300 mgl/ml) would cause less disruption of the blood-brain barrier than the relatively higher osmolality of iohexol (690 mOsm/kg at 300 mgl/ml). The degree of blood-brain barrier disruption was assessed qualitatively, by observing the degree of cortical staining with Evans' Blue dye, and quantitatively, by calculating the difference in uptake of 99mTc-pertechnetate between injected and noninjected hemispheres. Statistical analysis of the results showed that both iodixanol and iohexol had a significantly greater effect on blood-brain barrier disruption than did isotonic saline (0.005 greater than p greater than .001), but that the effect of iodixanol was not significantly different from that of iohexol with respect to either Evans' Blue staining (p greater than .05) or pertechnetate uptake (.75 less than p less than .90). Thus, the low-osmolality iodixanol has no significant advantage over iohexol in terms of blood-brain barrier disruption after experimental carotid angiography.