Purpose Modern multimodal analgesia has been shown to significantly reduce opioid use following total knee arthroplasty (TKA). This study was conducted to determine if changing TKA discharge opioid prescriptions from automatic to upon request resulted in more opioid free recoveries without compromising pain control. Methods Between December 2019 and August 2021, an orthopedic surgeon performed 144 primary unilateral TKAs; patients received the same multimodal analgesia protocol except for postoperative opioid prescribing. The first consecutively-treated cohort automatically received an opioid prescription following discharge (automatic group) and the second cohort received opioid prescriptions only upon request (upon request group). Opioid prescription data were derived from a prescription monitoring program and patient-reported outcomes (PROs) were collected preoperatively and at 2 and 12 weeks postoperatively. Results A higher percentage of the upon request group was opioid free 3 months after TKA compared with the automatic group (55.6% vs 4.3%, p < 0.0001) without compromising pain or function. Among opioid-naïve patients, 72% in the upon request group were opioid free after TKA compared with 5.4% in the automatic group. Opioid prescribing was not significantly reduced among opioid-experienced patients regardless of the pain protocol. Conclusion Requiring patients to request opioid prescriptions following TKA resulted in a higher rate of opioid free TKA, especially among opioid-naïve patients, without increasing pain compared with offering all patients an initial opioid prescription. Level of evidence Level III.