A comparison of stainable iron in simultaneously obtained aspirated smears and needle-biopsy specimens from 1,000 patients was undertaken. Significant differences occurred when iron was assessed as absent in the aspirated smear. In only 35% of the corresponding needle-biopsy specimens was iron absent. When only the aspirated smear was used, there was a significant overdiagnosis of iron deficiency. In general, iron tended to be less in the aspirated smear; however, correlations were better when iron stores were assessed as being present or increased in the aspirated smears, for stainable iron in the needle-biopsy specimen was always present in equal or greater amounts. Hemosiderotic smears (increased stainable iron) and needle-biopsy specimens (3+ -4+) correlated well. The aspirated smear and needle-biopsy are complementary procedures, and each has advantages. In the authors' experience, the needle-biopsy was preferable to the aspirated smear for evaluation of iron stores, particularly when iron stores were low or absent.