Observational studies primarily based on diet questionnaires or food records have reported that vegetarians can have a very low I intake. However, analytically ascertained data on the possible degree of I deficiency with this form of diet is lacking. Six healthy adult volunteers participated in the present controlled experimental diet study carried out in four separate 5 d diet periods. The study diets, normal, protein-rich, lactovegetarian, and repeat of the initial normal diet, were almost isoenergetic and contained no fish, sea food, iodized salt or processed foods fortified with I. During the last 48 h of each diet period two 24 h urine samples were obtained from each subject. I analyses were performed in the urine samples and in representative samples taken from all ingested diets. Urinary I excretion was significantly lower with the lactovegetarian diet (36·6 (sd 8·8) μg/d) than with the normal and the protein-rich diets (50·2 (sd 14·0) and 61·0 (sd 8·0) μg/d respectively). Accordingly, a markedly reduced I intake was confirmed analytically for the lactovegetarian diet (15·6 μg/d v. 35·2 and 44·5 μg/d respectively). Our results provide experimental confirmation of literature findings indicating that I supply is higher with non-vegetarian than with vegetarian diets. Specifically, the extremely low intake and urinary output of I as analytically determined for one exemplary vegetarian diet, demonstrate that dietary I may be limiting when strict forms of vegetarian dietary practices (no iodized salt, no I supplements) are followed. The present study is, therefore, the first diet-experiment-based pointer to the potential danger of I deficiency disorders due to strict forms of vegetarian nutrition, especially when fruits and vegetables grown in soils with low I levels are ingested.