We investigated the effects of supplement identification on exercise performance with caffeine supplementation. Forty‐two trained cyclists (age 37 ± 8 years, body mass [BM] 74.3 ± 8.4 kg, height 1.76 ± 0.06 m, maximum oxygen uptake 50.0 ± 6.8 mL/kg/min) performed a ~30 min cycling time‐trial 1 h following either 6 mg/kgBM caffeine (CAF) or placebo (PLA) supplementation and one control (CON) session without supplementation. Participants identified which supplement they believed they had ingested (“caffeine”, “placebo”, “don't know”) pre‐ and post‐exercise. Subsequently, participants were allocated to subgroups for analysis according to their identifications. Overall and subgroup analyses were performed using mixed‐model and magnitude‐based inference analyses. Caffeine improved performance vs PLA and CON (P ≤ 0.001). Correct pre‐ and post‐exercise identification of caffeine in CAF improved exercise performance (+4.8 and +6.5%) vs CON, with slightly greater relative increases than the overall effect of caffeine (+4.1%). Performance was not different between PLA and CON within subgroups (all P > 0.05), although there was a tendency toward improved performance when participants believed they had ingested caffeine post‐exercise (P = 0.06; 87% likely beneficial). Participants who correctly identified placebo in PLA showed possible harmful effects on performance compared to CON. Supplement identification appeared to influence exercise outcome and may be a source of bias in sports nutrition.