To document reasons for use of fresh frozen plasma (FFP) in dogs and determine variables that apparently triggered the decision to use FFP. Retrospective study. 74 dogs. Medical records of dogs that received FFP at a veterinary teaching hospital during a 3-month period were reviewed. The 74 dogs underwent 144 transfusion episodes (TE; a TE was defined as 1 day of transfusion therapy) and received 252 units (120 ml/unit) of FFP. Fresh frozen plasma was administered to provide coagulation factors (67 TE), albumin (91), alpha-macroglobulin (15), or immunoglobulins (19); for some TE, multiple clinical indications were identified. Variables that apparently triggered the decision to administer FFP included active hemorrhage with or without prolongation of coagulation times, low total plasma protein concentration, persistent vomiting associated with pancreatitis, and sepsis. Mean doses of FFP for each indication were between 8.5 and 9.4 ml/kg (3.9 and 4.3 ml/lb). Small dogs were generally given higher doses (mean dose, 13.9 ml/kg [6.3 ml/lb]) than large dogs (mean dose, 5.1 ml/kg [2.3 ml/lb]). Fifty (68%) dogs were alive at the time of discharge from the hospital. Results suggest that FFP plays an important role in the care of critically ill dogs. Because the supply of FFP is limited, guidelines for when administration of FFP may be clinically useful should be developed.