The use of rapid-acting insulin analogs as routes of administration other than IV has never been described for the treatment of dogs with diabetic ketoacidosis (DKA). This study aims to compare the efficacy and safety of a new protocol based on IM administration of insulin lispro with that of low-dose IV continuous rate infusion of regular insulin in the treatment of canine DKA. Client-owned dogs with naturally occurring DKA were included. Dogs treated with IM insulin lispro (Group L, n = 11) received 0.25 U/kg. The goal was to achieve a drop of at least 10% in blood glucose between 1 h and the next. If this goal was not achieved, the insulin dose was repeated hourly; otherwise, the insulin dose was not repeated up to a maximum of 3 h, after which the insulin dose was repeated anyway. When blood glucose was ≤250 mg/dL, the insulin dose was reduced to 0.125 U/kg IM every 3 h. Cases receiving IV continuous rate infusion of regular insulin (Group R, n = 13) were treated according to a previously published protocol. The median time to resolution of ketosis was significantly shorter in Group L (12 h; range, 4–27 h) compared to Group R (23 h; 10–46 h; P = 0.04). The median times to resolution of acidemia and ketoacidosis were 13 h (4–35 h) and 17.5 h (4–35 h) in Group L, and 22 h (9–80 h) and 23.5 h (10–80 h) in Group R, respectively. These differences were not significant (P = 0.06 and P = 0.09, respectively). The median length of hospitalization did not differ significantly between groups (P = 0.67). There were no differences in the frequency and severity of adverse events (hypoglycemia, hypokaliemia, and hypophosphatemia) between groups. The new protocol based on IM administration of insulin lispro preliminarily appears effective and safe for treatment of canine DKA.