The aim of this study was to investigate whether it was possible to reduce the time taken to complete a 4,000-m cycling time trial by misleading participants into believing they were racing against a previous trial, when in fact the power output was 2% greater. Nine trained male cyclists each completed four, 4,000-m time trials. The first trial was a habituation and the data from the second trial was used to form a baseline (BL). During trials three and four, participants raced against an avatar which they were informed represented their BL performance. However whilst one of these trials was an accurate (ACC) representation of BL, the power output in the other trial was set at 102% of BL and formed the deception condition (DEC). Oxygen uptake and respiratory exchange ratio were measured continuously and used to determine aerobic and anaerobic contributions to power output. There was a significant difference between trials for time to completion (F=15.3, p=0.00). Participants completed DEC more quickly than BL (90% CI: 2.1-10.1 s) and ACC (90% CI: 1.5-5.4 s); and completed ACC more quickly than BL (90% CI: 0.5-4.8 s). The difference in performance between DEC and ACC was attributable to a greater anaerobic contribution to power output at 90% of the total distance [(F=5.3, p=0.02), (90% CI: 4-37 W)]. The provision of surreptitiously augmented feedback derived from a previous performance reduces time taken for cyclists to accomplish a time trial of known duration. This suggests that cyclists operate with a metabolic reserve even during maximal time trials and that this reserve can be accessed following deception.