It is common for athletes striving to achieve maximal effort to exercise in the presence of a visible clock. It is implicitly assumed that calibration of the clock is normal (i.e. accurate). This study was designed to test the effect of secretly manipulating the clock calibration on maximal effort as measured by endurance times in cycle ergometry. Twelve subjects (6 male and 6 female) each undertook three identical rides to exhaustion on a cycle ergometer. In one the clock was normally calibrated, in another it was calibrated 10% faster, and in the third 10% slower. Tests were conducted double blind and in fully counterbalanced orders within gender. Clocked endurance times were recorded, and later converted to real times. Analysis of clocked times revealed no significant effects. Over all subjects, real endurance times showed a significant calibration effect, being on average 18.3% (73.4s) longer when the clock ran slow, compared to normal, and 20.5% (80.8s) longer when compared to fast. Because males exercised significantly longer than females, separate analyses reveal that the calibration effect was only significant in males, 27.7% (143.2s) and 29.7% (151.2s), respectively, and present but not significant in females, 1.3% (3.6s) and 3.8% (10.5s), respectively. These results suggest that, when deceived by a visible clock running slower than normal, times to exhaustion on the cycle ergometer were significantly longer in male subjects.