The incidence of metastatic carcinoma in bone marrow was studied by two different approaches in 220 consecutive autopsies of patients with carcinoma: histologic sections of aspirates from the iliac spine, as usually done in living patients; and gross and microscopic examination of the thoracolumbar spine. Ten percent of the autopsies showed bone marrow metastases by the aspiration technic as contrasted with 34.5% by the autopsy technic. Three cases had positive bone marrow aspirates in the absence of metastases demonstrable by routine autopsy. The combination of autopsy and aspiration technics yielded an incidence of metastatic marrow involvement of 35.9%. Of patients with marrow metastases at autopsy, 25% also had positive aspirates. This study establishes a norm for the comparison of the efficiency of discovery of metastatic carcinoma by different modalities of clinical bone marrow sampling.