Comparison of results of red cell mass (RCM) measurement by 51Cr and 125I methods in 119 patients showed virtual equivalence. Both methods have an acceptable coefficient of variation (CV) that is < 5%. The 125I method is simpler and much less expensive. Unrealistically narrow "normal ranges" for RCM are likely to lead to misdiagnosis of polycythemia vera. Upper normal limits of 39 mL/kg (males) and 32 mL/kg (females) are consistent with originally published data in normal persons; use of these limits as criteria would reduce the risk of misdiagnosis. No cases of "stress erythrocytosis" or Gaisbock Syndrome were encountered among the 119 cases reviewed.