Although donor livers with <30% large droplet macrovesicular steatosis (MaS) and/or small droplet MaS (irrespective of percentage) are considered safe to use, this consensus is based on variable definitions of MaS subtypes and/or without a reproducible scoring system. We analyzed 134 donor liver biopsies from allografts transplanted at University of California at San Francisco between 2000 and 2015 to determine whether large and/or small droplet MaS is a risk factor for poor outcomes. Large droplet MaS was defined as a fat droplet occupying greater than one half of an individual hepatocyte, with nuclear displacement, and scored as the percentage of total parenchymal area replaced by large fat droplets on ×40 magnification. Small droplet MaS was defined as 1 to several discrete fat droplets, each occupying less than one half of an individual hepatocyte, and scored as the percentage of remaining hepatocytes (ie, hepatocytes not occupied by large fat droplets) containing small fat droplets on ×200 magnification (ie, small droplet MaS is the percentage of “remaining hepatocytes” with small fat droplets, and “remaining hepatocytes” is defined as 100% minus percent large droplet MaS). Thus, total MaS equals the sum of large and small droplet MaS, which cannot exceed 100%. Electronic medical records were reviewed to determine outcomes. There was an increased risk for acute cellular rejection (hazard ratio=2.5, P=0.0108) and bile duct loss suggestive of chronic ductopenic rejection (hazard ratio=2.4, P=0.0130) in donor livers with ≥30% small droplet MaS. Large droplet MaS (up to 60%) was not associated with adverse outcomes. Patient survival was not adversely affected by steatosis. Excellent agreement on the estimation of large (weighted κ=0.682) and small droplet MaS (weighted κ=0.780) was achieved. Our approach to donor steatosis scoring can identify liver allograft recipients at increased risk for rejection and highlights the importance of distinguishing between small and large droplet MaS in this evaluation.