After preoperative conjunctival and periorbital antisepsis with povidone-iodine (PVP-I), the systemic absorption of iodine after cataract surgery was measured to evaluate the risk of thyroid side effects. Five different combinations of PVP-I alone or in combination with PVP-I-free antiseptics were applied to the conjunctiva and periorbital skin. An iodine-free product served as control. Iodide and creatinine in urine were analyzed before intervention and 24 and 48 h postoperatively. Depending on the concentration and application site, 0.3-4.5% of the total applied iodine or 3.6-45.4% of the free iodine were absorbed. The range of urine iodine excretion was between 11.7 and 71.0 mug iodine/g creatinine, depending on the PVP-I concentration and the site of application. The increase in iodine excretion was significant at 24 h postoperatively in trials receiving PVP-I both periorbitally and conjunctivally, depending of the concentration used. Because the iodine absorption is only slight and of doubtful clinical relevance, presurgical conjunctival antisepsis can be achieved with 1.25% PVP-I; so far clinically manifest anamnestic thyroid disorders are excluded. Presently, periorbital skin antisepsis with PVP-I cannot be recommended until data on thyroid metabolism in the population have been collected and evaluated, especially in a region currently or previously deficient in iodine.