Background: Social bonding, essential for health and survival in all social species, depends on mu-opioid signalling in non-human mammals. A growing neuroimaging and psychopharmacology literature also implicates mu-opioids in human social connectedness. Methods: To determine the role of mu-opioids for social connectedness in healthy humans, we conducted a preregistered (osf.io/x5wmq) multilevel random-effects meta-analysis of randomized double-blind placebo-controlled opioid antagonist studies. We included data from 8 publications and 2 unpublished projects, totalling 17 outcomes (N = 455) sourced from a final literature search in Web of Science, Scopus, PubMed and EMBASE on October 12, 2023, and through community contributions. All studies used naltrexone (25-100mg) to block the mu-opioid system and measured social connectedness by self-report. Results: Opioid antagonism slightly reduced feelings of social connectedness (Hedges’ g [95% CI) = -0.20 [-0.32, -0.07]. Results were highly consistent within and between studies (I2 = 23%). However, there was some indication of bias in favour of larger effects among smaller studies (Egger’s test: B = -2.16, SE = 0.93, z = -2.33, p = 0.02), and publication bias analysis indicated that the effect of naltrexone might be overestimated.Discussion: The results clearly demonstrate that intact mu-opioid signalling is not essential for experiencing social connectedness, as robust feelings of connectedness are evident even during full pharmacological mu-opioid blockade. Nevertheless, antagonism reduced measures of social connection, consistent with a modulatory role of mu-opioids for human social connectedness. The modest effect size relative to findings in non-human animals, could be related to differences in measurement (subjective human responses versus behavioural/motivation indices in animals), species specific neural mechanisms, or naltrexone effects on other opioid receptor subtypes. In sum, these results help explain how mu-opioid dysregulation and social disconnection can contribute to disability, and conversely – how social connection can buffer risk of ill health. Other: This work was supported by the European Research Council.