Purpose Fentanyl sublingual tablets (FST) are a potentially useful alternative to parenteral opioids such as subcutaneous morphine (SCM) to treat severe cancer pain episodes. No direct comparison between FST and SCM is available. The aim of this study was to test noninferiority of FST versus SCM during the first 30 min postadministration. Methods Patients receiving stable opioid therapy and experiencing a severe pain episode were randomly assigned to either 100 microg FST or 5 mg SCM in a double-blind, double-dummy trial. Average pain intensity (PI) assessed on a 0 to 10 numerical rating scale at 10, 20, and 30 min postadministration was the main end point. Analysis of covariance, adjusted by baseline PI, was the main analysis. The noninferiority margin (NIm) for the between-group difference was set at -0.6, that is, equal to one third of the minimum clinically important PI difference of two points. Results A total of 114 patients were randomly assigned to either FST (n = 58) or SCM (n = 56). One patient (in the FST group) withdrew consent before drug administration and was excluded from analysis. Baseline mean PIs were 7.5 in both groups; mean average PIs assessed at 10, 20, and 30 min postadministration were 5.0 and 4.5 for FST and SCM, respectively, with the 95% CI of the between-group difference including the NIm (-0.49; 95% CI, -1.10 to 0.09). Patients taking FST received a second drug dose after 30 min more frequently than did patients taking SCM (51% v 37 respectively; risk difference, -13 95% CI, -30% to 3. Both treatments were well tolerated, with average follow-up adverse event scores below the response of "A Little." Ninety-three percent of patients preferred the sublingual administration. Conclusion This trial did not show noninferiority of FST versus SCM within the chosen NIm. Both treatments were safe, and patients preferred the sublingual route of administration. FST provides analgesia with modest to moderate increased risk of lower efficacy compared with SCM.