Hypoalbuminemia and decreased colloid osmotic pressure (COP) are common in critically ill dogs. Hypoalbuminemia has been shown to negatively impact outcome. Treatment strategies to raise COP include the use of both natural and synthetic colloids. To date there is no comprehensive evaluation of human albumin (HA) use in dogs. The purpose of this retrospective study was to evaluate the use of 25% HA in critically ill dogs. The medical records of all dogs receiving HA between January 2003 and March 2004 were reviewed using a standardized data sheet for signalment, diagnoses, Survival Prediction Index (SPI), dose of HA administered, amount of crystalloid fluid administered prior to HA administration, length of hospitalization (LOH), complications, and outcome. Additionally pre- and post-HA administration values for serum albumin (SA), blood COP, and total solids (TS) were evaluated. Thirty-seven dogs received HA. Thirty of the 37 dogs (81%) underwent surgery and 13 of 37 (39%) dogs were diagnosed with septic peritonitis. Other diagnoses included trauma, wounds, neoplasia, gastric ulcer, and pancreatitis. The median SPI of all dogs was 0.53 (range 0.06–0.82). Median dose of HA used was 1.5 g/kg (range 0.38–3.64 g/kg). Median dose of crystalloid fluids administered prior to HA administration was 6.23 ml/kg/hr (range 1.2–45.4 ml/kg/hr). Median LOH was 5 days (range 1–16 days). Mild complications were encountered in 9 of 37 (24%) dogs and included prolonged clotting times (n=3), increased breathing effort (n=3), vomiting (n=2), and fever (n=1). Of the 37 dogs in the study, 21 (57%) survived, 10 (27%) died, and 6 (16%) were euthanized. After HA administration, median SA increased from 1.5 g/dl (range 1.0–2.1 g/dl) to 2.8 g/dl (range 1.5–4.4 g/dl) (p<0.001), median COP increased from 11 mmHg (range 6.8–13.8 mmHg) to 14.4 mmHg (range 11–26.9 mmHg) (p<0.001), and median TS increased from 2.9 g/dl (range 0.9–4.6 g/dl) to 4.0 (range 2.8–6.2) (p<0.001). When compared to non-survivors, survivors had higher post-HA administration SA (p=0.008), higher net COP increase (p=0.032), and received higher dosages of HA (p=0.046). Pre-treatment SA, COP, or the development of complications during administration did not impact LOH or outcome. In this retrospective study, HA administration effectively increased SA, TS and COP in critically ill dogs and was associated with relatively few complications. Future studies are warranted to further evaluate the potential benefits of HA administration.