The iodine intake level in a population is determined in cross-sectional studies. Urinary iodine varies considerably and the reliability of studies of iodine nutrition and the number of samples needed is unsettled. We performed a longitudinal study of sixteen healthy men living in an area of mild to moderate iodine deficiency. Iodine and creatinine concentrations were measured in spot urine samples collected monthly for 13 months. From these data we calculated the number of urine samples needed to determine the iodine excretion level for crude urinary iodine and for 24 h iodine excretion estimated from age- and gender-specific creatinine excretions. We found that mean urinary iodine excretion varied from 30 to 87 μg/l (31 to 91 μg/24 h). Sample iodine varied from 10 to 260 μg/l (20 to 161 μg/24 h). Crude urinary iodine varied more than estimated 24 h iodine excretion (population standard deviation 32 v. 26; individual standard deviation 29 v. 21; Bartlett's test, P < 0·01 for both). The number of spot urine samples needed to estimate the iodine level in a population with 95 % confidence within a precision range of ± 10 % was about 125 (100 when using estimated 24 h iodine excretions), and within a precision range of ± 5 % was about 500 (400). A precision range of ± 20 % in an individual required twelve urine samples or more (seven when using estimated 24 h iodine excretions). In conclusion, estimating population iodine excretion requires 100–500 spot urine samples for each group or subgroup. Less than ten urine samples in an individual may be misleading.