Serum ESF titers were measured in 42 polycythemic patients using the fetal mouse liver cell bioassay. ESF titers in patients with secondary polycythemia differed significantly from those in patients with polycythemia vera (p less than 0.0001). Among the 21 patients with secondary polycythemia, 1 patient had an ESF titer less than 10 mU/ml (the lower limit of sensitivity) and 20 had ESF titers that ranged between 11 and 112 mU/ml, with a mean titer of 56 mU/ml. Among the 21 patients with polycythemia vera, 13 patients had ESF titers less than 10 mU/ml and 8 had ESF titers ranging between 12 and 55 mU/ml, with a mean titer of 26 mU/ml. The mean hemoglobin concentration in the 8 patients with ESF titers greater than 10 mU/ml was significantly below that in the 13 polycythemia vera patients with ESF titers less than 10 mU/ml (p less than 0.03). If ESF titers less than 10 mU/ml had been indicative of polycythemia vera and ESF titers greater than 10 mU/ml had been indicative of secondary polycythemia in patients with hemoglobin concentrations greater than 17.7 g/dl, but not indicative of either condition in patients with hemoglobin concentrations less than 17.7 g/dl, 71.5% of the polycythemic patients in this study would have been diagnosed correctly, 9.5% incorrectly, and in the 19% the diagnosis would have remained uncertain. It was concluded that measurement of serum ESF titers using this in vitro bioassay can be of clinical importance in differentiating between polycythemia vera and secondary polycythemia.