In 111 patients, the amount of stainable iron was compared in bone marrow smears and marrow needle biopsies. The correlation between the two methods was good (r = 0,73, p<0,001), but with a higher frequency of false negative results in the marrow biopsies (5,4%) than the marrow smears (1,8%). As both methods may occasionally fail, it may be desirable to perform iron staining in both smears and biopsies. If it is necessary to choose one of the methods, we would recommend bone marrow smears for the routine determination of stainable marrow iron stores.