The relative risk of adverse drug reactions of ionic versus non-ionic contrast media injected i.v. were compared for different types of trials using odds-ratio. The absolute and relative risk found in large post-marketing trials were compared with that found in the iohexol pre-registration trials. The absolute risks were 2 to 10 times higher in the pre-registration trials compared to the post-marketing surveillances. The relative risk for all adverse drug reactions was 3 to 6 times higher for ionic vs. non-ionic media and independent of pre- or post-registration studies. The odds-ratio seems to be a feasible method of comparing the relative risk of adverse reactions in various trials.