BACKGROUND AND OBJECTIVES: The detection of recombinant human erythropoietin (r-HuEPO) abuse by athletes remains problematic. The main aim of this study was to demonstrate that the five indirect markers of altered erythropoiesis identified in our earlier work were reliable evidence of current or recently discontinued r-HuEPO use. A subsidiary aim was to refine weightings of the five markers in the initial model using a much larger data set than in the pilot study. A final aim was to verify that the hematologic response to r-HuEPO did not differ between Caucasian and Asiatic subjects. DESIGN AND METHODS: Recreational athletes resident in Sydney, Australia (Sydney, n = 49; 16 women, 33 men) or Beijing, China (Beijing, n=24; 12 women, 12 men) were randomly assigned to r-HuEPO or placebo groups prior to a 25 day administration phase. Injections of r-HuEPO (or saline) were administered double-blind at a dose of 50 IU/kg three times per week, with oral iron (105 mg) or placebo supplements taken daily by all subjects. Blood profiles were monitored during and for 4 weeks after drug administration for hematocrit (Hct), reticulocyte hematocrit (RetHct), percent macrocytes (%Macro), serum erythropoietin (EPO) and soluble transferrin receptor (sTfr), since we had previously shown that these five variables were indicative of r-HuEPO use. RESULTS. The changes in Hct, RetHct, %Macro, EPO and sTfr in the Sydney trial were qualitatively very similar to the changes noted in our previous administration trial involving recreational athletes of similar genetic origin. Statistical models developed from Fisher's discriminant analysis were able to categorize the user and placebo groups correctly. The same hematologic response was demonstrated in Beijing athletes also administered r-HuEPO. INTERPRETATION AND CONCLUSIONS: This paper confirms that r-HuEPO administration causes a predictable and reproducible hematologic response. These markers are disturbed both during and for several weeks following r-HuEPO administration. This work establishes an indirect blood test which offers a useful means of detecting and deterring r-HuEPO abuse. Ethnicity did not influence the markers identified as being able to detect athletes who abuse r-HuEPO.