BACKGROUND: Although interscalene nerve block is the standard for shoulder analgesia, the risk of hemidiaphragmatic paralysis restricts its use in patients with compromised pulmonary function. We hypothesized that a novel subparaneural upper trunk block would provide noninferior postoperative analgesia but superior diaphragmatic sparing effect compared to interscalene block. METHODS: This randomized controlled trial comprised 96 patients who underwent arthroscopic shoulder surgery under either subparaneural upper trunk block (5 mL of 0.5% ropivacaine) or interscalene block (15 mL of 0.5% ropivacaine), followed by supraclavicular nerve block (5 mL of 0.5% ropivacaine). General anesthesia was standardized. The coprimary outcomes were (1) recovery room resting pain score at 30 minutes, measured on an 11-point numerical rating scale, with a prespecified noninferiority margin of 1 point and (2) the incidence of hemidiaphragmatic paralysis, diagnosed using ultrasound. Among secondary outcomes, resting pain scores were assessed with numerical rating scale at 4, 8, and 24 hours postoperatively. RESULTS: Recovery room resting pain scores at 30 minutes were 0 (0-1) in the subparaneural upper trunk group versus 0 (0-0) in the interscalene group, with a median difference of 0 (95% CI, 0-0); the upper 95% CI limit was lower than the prespecified noninferiority margin (noninferiority P < .001). Hemidiaphragmatic paralysis was observed in 16.7% of patients in the subparaneural upper trunk group versus 100% of those in the interscalene group (RR, 0.17; 95% CI, 0.09-0.31; P < .001), with complete paralysis occurring in 6.3% and 93.7% of patients, respectively. In this study, any reported differences in pain scores at 4, 8, and 24 hours postoperatively were not clinically important. CONCLUSIONS: The subparaneural upper trunk block compared to interscalene block provided noninferior analgesia at 30 minutes in the recovery room after arthroscopic shoulder surgery but resulted in less hemidiaphragmatic paralysis.