Purpose: The primary objective was to assess the performance benefits of caffeine (CAF) supplementation in habitual users. Importantly, this investigation was designed to account for the potential confounding effects of CAF withdrawal (CAFW), which are inherent and common in previous work. Methods: Ten CAF-consuming (394 [146] mg·d−1) recreational cyclists (age 39.1 [14.9] y; maximum oxygen consumption 54.2 [6.2] mL·kg–1·min–1) completed four 10-km time trials (TTs) on a cycle ergometer. On each trial day, 8 hours before reporting to the laboratory, subjects consumed 1.5 mg·kg–1 CAF to prevent withdrawal (no withdrawal [N]) or a placebo (PLA; withdrawal [W]). Then, 1 hour prior to exercise, they received either 6 mg·kg–1 CAF or PLA. These protocols were repeated 4 times, employing all combinations of N/W and CAF/PLA. Results: CAFW did not impair TT power output (PLAW vs PLAN P = .13). However, preexercise CAF only improved TT performance when compared to PLA in the W condition (CAFN vs PLAW P = .008, CAFW vs PLAW P = .04), not when W was mitigated (PLAN vs CAFN P = .33). Conclusions: These data indicate that preexercise CAF only improves recreational cycling performance when compared to bouts preceded by CAF abstinence, suggesting that habitual users may not benefit from 6 mg·kg–1 of CAF and that previous work may have overstated the value of CAF supplementation for habitual users. Future work should examine higher doses of CAF for habitual users.