OBJECTIVE To test the hypothesis that iodinated contrast media may induce an elevation in serum potassium level. METHODS Protocol A: After intravenous infusion of contrast media into six rabbits, alterations of potassium ion concentrations were measured. Protocol B: Fresh rabbit blood was mixed in vitro with contrast media, and the fluctuations in potassium were monitored over a 30-minute period. Protocol C: Similar to protocol B, except that blood from humans with no reaction to contrast media was used. RESULTS For protocol A, blood potassium levels increased above baseline levels. The elevations were statistically significant (P < .05). For protocol B, diatrizoate and ioxaglate caused a gradual increase in blood potassium levels, but iopamidol did not. In protocol C, all three contrast media caused statistically significant elevation in potassium levels. The release of potassium was statistically significant at 5 minutes (P < .05 for diatrizoate and ioxaglate, and P < .01 for iopamidol). The mean release rates (+/- standard deviation) by means of linear regression analysis were 0.0190 mmol/min +/- 0.0112 with diatrizoate, 0.0159 mmol/min +/- 0.0057 with iopamidol, and 0.0088 mmol/min +/- 0.0033 with ioxaglate. CONCLUSIONS Iodinated contrast media increase blood potassium levels causing release of potassium into intravascular spaces. This potassium release may play some role in contrast medium-induced adverse reactions.