What it is
Oxytocin is a naturally occurring 9-amino-acid human hormone, released by the posterior pituitary, with well-established roles in labor, milk letdown, and social/pair-bonding behavior in animal models. A synthetic version has been an FDA-approved drug for decades, sold generically and under the brand name Pitocin, given intravenously or intramuscularly to induce or augment labor and to control postpartum bleeding.
A separate, much newer research and commercial story concerns intranasal oxytocin, studied for its effects on social cognition, trust, anxiety, and psychiatric conditions including autism spectrum disorder and schizophrenia. This intranasal use has no FDA approval, and it should not be confused with the approved obstetric drug: different route, different population, different evidence base entirely.
The approved use versus the circulating one
The obstetric use is old, well-studied, and dosed precisely by IV infusion titrated to uterine response, under labor and delivery monitoring.
The intranasal, psychiatric/social-cognition use is a different story. A 2026 meta-analysis pooling 42 double-blind randomized controlled trials of intranasal oxytocin versus placebo across autism spectrum disorder, schizophrenia-spectrum disorders, substance use disorders and other mental disorders (total n=1,922) found a small, non-significant overall effect (g=0.17, 95% CI -0.05 to 0.38) with substantial heterogeneity between trials; removing two outlier studies eliminated the heterogeneity and reduced the pooled effect further (g=0.05, 95% CI -0.03 to 0.12), essentially no effect. A significant, small benefit emerged only in the schizophrenia-spectrum subgroup (g=0.12).
Specifically for autism, an earlier systematic review and meta-analysis of 10 randomized controlled trials in adults found no significant efficacy of intranasal oxytocin for anxiety, repetitive behavior, social function or overall severity.
What is not established
Any general benefit of intranasal oxytocin for social cognition, bonding, anxiety or mood in a healthy person taking it outside a clinical trial. The largest recent meta-analysis pooling trials across several psychiatric conditions found an effect close to zero once outlier studies were removed.
Optimal dose, frequency or duration for any of the unapproved intranasal uses; the 2026 meta-analysis specifically found no significant moderation by dose or number of administrations across the pooled trials, meaning the field has not established a dose-response relationship even within its own trial data.