Time to task failure during high-intensity, exhaustive cycling exercise might coincide with a severe degree of peripheral fatigue, which is never voluntarily exceeded by the exercising human (i.e. a “critical threshold”). However, data showed that it is possible to improve performance time during cycling time trial (TT), by misleading participants into believing they are racing against a virtual avatar reproducing a previous TT when, in fact, power output was greater (i.e. deceptive TT). The neuromuscular consequences underpinning this performance improvement remains to be determined. Therefore, the present study aimed to investigate the influence of deceptive TTs on end-exercise peripheral and central fatigue.Ten participants performed, on separate days, one maximal 5km TT (5KCTRL) and three deceptive 5km TTs, during which they had to strictly adjust their pace to match the one of a virtual pacer. Whilst participants were informed that the virtual pacer reproduced the pattern of power output measured during 5KCTRL, which was true in only one session (5K100%), power output was in fact 5% (5K105%) and 10% (5K110%) greater in two other sessions. Maximal voluntary contraction (MVC) and peripheral fatigue were quantified via changes in pre- to post-exercise isometric quadriceps force and potentiated quadriceps twitch force (QTW) as evoked by supramaximal electrical stimulation of the femoral nerve, respectively. Central fatigue was estimated via changes in pre- to post-exercise quadriceps voluntary activation (VA). Power output (229±64W vs 219±61W, P<0.001) and performance time (9.03±1.2min vs 9.16±1.2min, P<0.001) were improved by 5±2% and 2±1% in 5K105% compared to 5K100%, respectively. These improvements were associated with a 14±6% and 28±9% greater reduction in MVC (-41±14% vs -36±14%, P<0.01) and VA (-19±10% vs -15±9%, P<0.01) but unchanged QTW (-51±8% vs -50±9%, P=0.55) in 5K105% vs 5K100%, respectively. During 5K110%, participants failed (P>0.25) to improve power output (223±60W) and performance time (9.12±1.1min) compared to 5K100%. Following 5K110%, reduction in MVC (-40±14%), VA (-16±9%) and QTW (-51±8%) were not different compared to 5K100% (P>0.23).Our data showed that performance improvement during a deceptive TT is limited to small increments in power output and occurred at the expense of a centrally-mediated reduction in the muscle force generating capacity. This rise in central fatigue might assure that muscle activation does not surpass a level that would cause a further accumulation of peripheral fatigue beyond an invariable level.