The sensitivity of the adaptive model of the Athlete Biological Passport (ABP) and reticulocyte percentage (ret%) in detection of recombinant human erythropoietin (rHuEPO) misuse was evaluated using both a long-term normal dose and a brief high dose treatment regime. Sixteen subjects received either 65 IU rHuEPO × kg every second day for two weeks (normal-dose), 390 IU rHuEPO × kg on three consecutive days (high-dose), or frequent placebo treatment for 13 days in a randomized, placebo-controlled, double-blind crossover design. Blood variables were measured 4, 11, and 25 days following treatment initiation. The ABP based on haemoglobin concentration ([Hb]) and OFF-hr score ([Hb] - 60 × √ret%) yielded atypical profiles following both normal-dose and high-dose treatment (0 %, 31 %, 13 % vs. 21 %, 33 %, 20 % at days 4, 11, and 25 after normal and high dose, respectively). Including ret% as a stand-alone marker for atypical blood profiles increased (P < 0.05) the sensitivity of the adaptive model at day 11 to 63 % and 67 % for normal-dose and high-dose rHuEPO administration, respectively. In conclusion, ~30 % of subjects injecting a normal-dose rHuEPO for two weeks or a high-dose rHuEPO for three days will present an atypical ABP profile. Including ret% as a stand-alone parameter improves the sensitivity two-fold. Copyright © 2015 John Wiley & Sons, Ltd.