The purpose of this study was to analyze the pacing strategy and its affective consequences during self-paced cycling time trials (TT) performed at different severity of hypoxia. Eight competitive cyclists performed five 30 min self-paced TTs at their best performance in the following conditions: 1) normobaric normoxia (NNSL); 2) normobaric hypoxia under two simulated altitudes: 2000 m (NH2000) and 3500 m (NH3500) and 3) normobaric hypoxia but the cyclists were deceived and thought to be at sea level for 2000 m (DecNH2000) and 3500 m (DecNH3500). Power Output (PO), oxygen uptake (VO2), and blood lactate concentration ([La]) were recorded to assess exercise intensity and physiological adaptations. The rate of perceived exertion (RPE) and pleasure were measured with a CR10 Borg scale to evaluate the affective load (AL). PO and VO2 decreased with the severity of hypoxia but no significantly difference on performance was measured between deceived and real conditions, except for pacing strategy. The started intensity depends on the exercise expectations, but PO was rapidly adjusted with the physiological constraints and the rate of increase of RPE. Finally, AL did not reach maximal values so that the athletes sustained a physiological and emotional reserve to perform a final spurt.