This paper presents engine dynamometer testing and modeling analysis of ethanol compared to gasoline at part load conditions where the engine was not knock-limited with either fuel. The purpose of this work was to confirm the efficiency improvement for ethanol reported in published papers, and to quantify the components of the improvement.Testing comparing E85 to E0 gasoline was conducted in an alternating back-to-back manner with multiple data points for each fuel to establish high confidence in the measured results. Approximately 4% relative improvement in brake thermal efficiency (BTE) was measured at three speed-load points.Effects on BTE due to pumping work and emissions were quantified based on the measured engine data, and accounted for only a small portion of the difference. Approximately half of the improvement was accounted for by the fact that the heat of vaporization (HoV) of the fuel detracts from the heat release measured in the combustion bomb used in the determination of heating value, but does not detract from the heat released during combustion in the engine. Engine modeling indicated that the remaining difference in BTE is due to lower burned gas temperatures and consequently lower heat transfer losses. The lower temperatures are due to greater charge cooling and to lower adiabatic flame temperature.CO₂ emissions at part load are reduced about 7% for ethanol compared to gasoline. Approximately 4% CO₂ benefit is due to improved thermal efficiency, and about 3% is due to the increased hydrogen-to-carbon ratio (H/C) of ethanol.