ObjectiveTo determine the dose and cardiopulmonary effects of propofol alone or with midazolam for induction of anesthesia in American Society of Anesthesiologists status ≥III dogs requiring emergency abdominal surgery. Study designProspective, randomized, blinded, clinical trial. AnimalsA total of 19 client-owned dogs. MethodsDogs were sedated with fentanyl (2 μg kg–1) intravenously (IV) for instrumentation for measurement of heart rate, arterial blood pressure, cardiac index, systemic vascular resistance index, arterial blood gases, respiratory rate and rectal temperature. After additional IV fentanyl (3 μg kg–1), the quality of sedation was scored and cardiopulmonary variables recorded. Induction of anesthesia was with IV propofol (1 mg kg–1) and saline (0.06 mL kg–1; group PS; nine dogs) or midazolam (0.3 mg kg–1; group PM; 10 dogs), with additional propofol (0.25 mg kg–1) IV every 6 seconds until endotracheal intubation. Induction/intubation quality was scored, and anesthesia was maintained with isoflurane. Variables were recorded for 5 minutes with the dog in lateral recumbency, breathing spontaneously, and then in dorsal recumbency with mechanical ventilation for the next 15 minutes. A general linear mixed model was used with post hoc analysis for multiple comparisons between groups (p < 0.05). ResultsThere were no differences in group demographics, temperature and cardiopulmonary variables between groups or within groups before or after induction. The propofol doses for induction of anesthesia were significantly different between groups, 1.9 ± 0.5 and 1.1 ± 0.5 mg kg–1 for groups PS and PM, respectively, and the induction/intubation score was significantly better for group PM. Conclusions and clinical relevanceMidazolam co-induction reduced the propofol induction dose and improved the quality of induction in critically ill dogs without an improvement in cardiopulmonary variables, when compared with a higher dose of propofol alone.