There is still no consensus regarding the indications of long-acting injectable antipsychotics (LAIs) in early psychosis (EP). This umbrella review synthesizes findings from systematic reviews and meta-analyses on the risk-benefit balance of LAIs in EP. Eligible systematic reviews and meta-analyses on LAIs in EP were identified by a MEDLINE search from inception until June 2024. Data were synthesized narratively. Seven systematic reviews and four meta-analyses published from 2007 to 2024 were identified. They included 220 to 14,313 participants recruited in 58 primary unique randomized controlled trials or observational studies. All LAIs were considered in most reviews. Inclusion criteria and diagnoses differed widely across the selected reviews. The reviews and meta-analyses consistently showed a positive impact of LAIs on symptomatic outcome in people with EP, although there was no consensus on whether LAIs outperformed oral anti-psychotics (OAPs). Most reported a greater reduction of treatment discontinuation due to inefficacy or nonadherence with LAIs vs. OAPs, although meta-analyses found no difference between LAIs vs. OAPs regarding all-cause discontinuation. Findings regarding relapse prevention were inconclusive. Similar rates of metabolic adverse drug reaction and potentially lower rates of extrapyramidal symptoms were observed with LAIs vs. OAPs. LAIs should be considered according to users' preferences and could be particularly useful for people with a poor medication adherence risk profile. Further high-quality observational studies in real-life prescribing conditions are needed to support robust recommendations for clinical practice regarding indications of LAIs in EP.< Réduire