Vagus nerve stimulation (VNS) has shown promise as an adjuvant treatment for posttraumatic stress disorder (PTSD), as it enhances fear extinction and reduces anxiety symptoms in multiple rat models of this condition. Yet, identification of the optimal stimulation paradigm is needed to facilitate clinical translation of this potential therapy. Using an extinction-resistant rat model of PTSD, we tested whether varying VNS intensity and duration could maximize extinction learning while minimizing the total amount of stimulation. We confirmed that sham rats failed to extinguish after a week of extinction training. Delivery of the standard LONG VNS trains (30 s) at 0.4 mA enhanced extinction and reduced anxiety but did not prevent fear return. Increasing the intensity of LONG VNS trains to 0.8 mA prevented fear return and attenuated anxiety symptoms. Interestingly, delivering 1, 4 or 16 SHORT VNS bursts (0.5 s) at 0.8 mA during each cue presentation in extinction training also enhanced extinction. LONG VNS trains or multiple SHORT VNS bursts at 0.8 mA attenuated fear renewal and reinstatement, promoted extinction generalization and reduced generalized anxiety. Delivering 16 SHORT VNS bursts also facilitated extinction in fewer trials. This study provides the first evidence that brief bursts of VNS can enhance extinction training, reduce relapse and support symptom remission using much less VNS than previous protocols. These findings suggest that VNS parameters can be adjusted in order to minimize total charge delivery and maximize therapeutic effectiveness.