We estimate the effects of the 1924 introduction of iodized salt in the U.S. by exploiting pre-1924 geographic variation in iodine deficiency. Iodized salt reduced infant mortality by 1.1 deaths per 1,000 births (1.6%) for counties at the 75th percentile of iodine deficiency relative to the 25th percentile. These effects are concentrated in urban counties—where iodized salt was disproportionately available—explaining 1/3 of the decline in the urban-rural infant mortality gap in the 1920s. We show that the long-term effects on labor market outcomes are consistently large among urban-born individuals, reconciling conflicting results in the literature.