The absorption of iron from three different iron tablets - rapidly-disintegrating ferrous sulphate and ferrous carbonate tablets and slow-release ferrous sulphate tablets - was studied in healthy subjects with a serum iron technique and by whole-body counter measurements. A solution of ferrous sulphate was used as a reference. There were no differences in the absorption from the ferrous sulphate preparations, but the ferrous carbonate tablets were less well absorbed. Good correlation was found between the maximal serum iron response and the total absorption of iron and it was concluded that serum iron studies may be used for semiquantitative measurements of iron absorption in comparative studies on different iron preparations.