Purpose: To evaluate the influence of warm-up exercise intensity and subsequent recovery on intense endurance performance, selected blood variables, and the oxygen-uptake (VO2) response. Methods: Twelve highly trained male cyclists (VO2max 72.4 +/- 8.0 mL.min(-1).kg(-1), incremental-test peak power output (iPPO) 432 +/- 31 W; mean +/- SD) performed 3 warm-up strategies lasting 20 min before a 4-min maximal-performance test (PT). Strategies consisted of moderate-intensity exercise (50%iPPO) followed by 6 min of recovery (MOD6) or progressive high-intensity exercise (10-100%iPPO and 2 x 20-s sprints) followed by recovery for 6 min (HI6) or 20 min (HI20). Results: Before PT venous pH was lower (P < .001) in HI6 (7.27 +/- 0.05) than in HI20 (7.34 +/- 0.04) and MOD6 (7.35 +/- 0.03). At the same time, differences (P < .001) existed for venous lactate in HI6 (8.2 +/- 2.0 mmol/L), HI20 (5.1 +/- 1.7 mmol/L), and MOD6 (1.4 +/- 0.4 mmol/L), as well as for venous bicarbonate in HI6 (19.3 +/- 2.6 mmol/L), HI20 (22.6 +/- 2.3 mmol/L), and MOD6 (26.0 +/- 1.4 mmol/L). Mean power in PT in HI6 (402 38 W) tended to be lower (P = .11) than in HI20 (409 +/- 34W) and was lower (P = .007) than in MOD6 (416 +/- 32W). Total VO2 (15-120 s in PT) was higher in H16 (8.18 +/- 0.86 L) than in H120 (7.85 +/- 0.82 L, P = .008) and MOD6 (7.90 +/- 0.74 L, P = .012). Conclusions: Warm-up exercise including race-pace and sprint intervals combined with short recovery can reduce subsequent performance in a 4-min maximal test in highly trained cyclists. Thus, a reduced time at high exercise intensity, a reduced intensity in the warm-up, or an extension of the recovery period after an intense warm-up is advocated.