BACKGROUNDPeripheral regional anaesthesia is frequently used for upper extremity surgery. To prolong the duration of analgesia, adjuvants can be added to single-injection local anaesthetics. Despite attempts to compare several adjuvants in pairwise meta-analyses, a comprehensive comparison is still missing.OBJECTIVEThe objective of this network meta-analysis was to determine the effectiveness of adjuvants in upper extremity peripheral nerve blocks.DESIGNA systematic review of randomised controlled trials with network meta-analyses.DATA SOURCESA literature search in Embase, CENTRAL, MEDLINE and Web of Science was performed up to March 2023.ELIGIBILITY CRITERIARandomised trials comparing different adjuvants injected perineurally in peripheral upper extremity nerve blocks were eligible. Frequentist network meta-analysis was conducted using a random effects model with physiological saline as the comparator. The primary endpoint was the ratio of means (ROM) of the duration of analgesia.RESULTSThe review included 242 randomised controlled trials with a total of 17 391 patients. Twenty-eight adjuvants were compared in the largest networks. Most network estimations consisted of a high proportion of direct evidence. Fourteen adjuvants increased the duration of analgesia significantly by the following factors, ROM [95% confidence interval (CI)]: dexamethasone 1.95 (1.79 to 2.13), buprenorphine 1.83 (1.51 to 2.24), butorphanol 1.84 (1.41 to 2.39), potassium chloride 1.89 (1.15 to 3.11), dexmedetomidine 1.70 (1.59 to 1.81), sufentanil 1.70 (1.27 to 2.29), ketorolac 1.68 (1.24 to 2.27), midazolam 1.55 (1.24 to 1.94), tramadol 1.52 (1.32 to 1.75), nalbuphine 1.50 (1.30 to 1.72), morphine 1.43 (1.09 to 1.88), magnesium sulfate 1.42 (1.20 to 1.67), clonidine 1.36 (1.24 to 1.50) and fentanyl 1.23 (1.08 to 1.40). Inconsistency in network meta-analysis was substantial. Overall side effect rates were low with all adjuvants.CONCLUSIONThe best interventions to prolong the duration of analgesia were dexamethasone, followed by dexmedetomidine, opioids, electrolytes, ketorolac and midazolam. There are general concerns about the quality of underlying studies and the risk of publication bias.TRIAL REGISTRATIONPROSPERO 2018 CRD42018115722