Background: Although cyclists often compete at altitude, the effect of altitude on gross efficiency (GE) remains inconclusive. Purpose: To investigate the effect of altitude on GE at the same relative exercise intensity and at the same absolute power output (PO) and to determine the effect of altitude on the change in GE during high-intensity exercise. Methods: Twenty-one trained men performed 3 maximal incremental tests and 5 GE tests at sea level, 1500 m, and 2500 m of acute simulated altitude. The GE tests at altitude were performed once at the same relative exercise intensity and once at the same absolute PO as at sea level. Results: Altitude resulted in an unclear effect at 1500 m (−3.8%; ±3.3% [90% confidence limit]) and most likely negative effect at 2500 m (−6.3%; ±1.7%) on pre-GE, when determined at the same relative exercise intensity. When pre-GE was determined at the same absolute PO, unclear differences in GE were found (−1.5%; ±2.6% at 1500 m; −1.7%; ±2.4% at 2500 m). The effect of altitude on the decrease in GE during high-intensity exercise was unclear when determined at the same relative exercise intensity (−0.4%; ±2.8% at 1500 m; −0.7%; ±1.9% at 2500 m). When GE was determined at the same absolute PO, altitude resulted in a substantially smaller decrease in GE (2.8%; ±2.4% at 1500 m; 5.5%; ±2.9% at 2500 m). Conclusion: The lower GE found at altitude when exercise is performed at the same relative exercise intensity is mainly caused by the lower PO at which cyclists exercise.