OBJECTIVES: Patients labelled with penicillin allergy (PenA) often receive broader spectrum antibiotics, associated with antimicrobial resistance and poorer outcomes. However, ∼95% of patients are likely mis-labelled. Whilst de-labelling programmes are gaining momentum, they have been restricted to a few countries. Here, we address the global prevalence of PenA, to inform the wider potential impact for de-labelling programmes. METHODS: We conducted a systematic review and meta-analysis including all studies on adult PenA prevalence between January 2003 and June 2023. Data on PenA prevalence, allergy recording methods, healthcare setting, and country income were extracted. This study is registered on PROSPERO (CRD42023437718). RESULTS: 174 studies from 28 countries were included (18,352 screened). Global PenA prevalence was 9·4% (95% CI 8·4-10·4%). 92% of peer-reviewed publications were from high-income countries(HICs), with 72% from the UK, USA or Australia. HICs had higher PenA prevalence 9·9% (95% CI 8·8-11·0%), compared to middle-income countries (MICs), 4·4% (95% CI 2·8-6·2%), p<
0.0001. Primary care data was seldom reported (16% of studies), and the method of allergy recording significantly influenced reported prevalence. CONCLUSIONS: Studies reporting PenA prevalence are skewed towards HICs and secondary care, with little data from Africa, most of Asia and South America. This highlights an unmet need to broaden epidemiological analysis in under-represented regions.