During construction work (2017–2019), an increase in Aspergillus flavusinfections was noted among pediatric patients, the majority of whom were receivingamphotericin B prophylaxis. Microsatellite genotyping was used to characterize theoutbreak. A total of 153 A. flavus isolates of clinical and environmental origin wereincluded. Clinical isolates included 140 from 119 patients. Eight patients were outbreak-related patients, whereas 111 were outbreak-unrelated patients from Danish hospitals(1994–2023). We further included four control strains. Nine A. flavus isolates were fromsubsequent air sampling in the outbreak ward (2022–2023). Typing followed Rudramurthy et al.(S. M. Rudramurthy, H. A. de Valk, A. Chakrabarti, J. Meis, and C. H. W. Klaassen, PLoS One 6:e16086, 2011, https://doi.org/10.1371/journal.pone.0016086). Minimumspanning tree (MST) and discriminant analysis of principal components (DAPC) wereused for cluster analysis. DAPC analysis placed all 153 isolates in five clusters. Microsatellite marker pattern was clearly distinct for one cluster compared to the others.The same cluster was observed in an MST. This cluster included all outbreak isolates,air-sample isolates, and additional patient isolates from the outbreak hospital, previouslyundisclosed as outbreak related. The highest air prevalence of A. flavus was found intwo technical risers of the outbreak ward, which were then sealed. Follow-up air sampleswere negative for A. flavus. Microsatellite typing defined the outbreak as nosocomialand facilitated the identification of an in-hospital source. Six months of follow-upair sampling was without A. flavus. Outbreak-related/non-related isolates were easilydistinguished with DAPC and MST, as the outbreak clone’s distinct marker pattern wasdelineated in both statistical analyses. Thus, it could be a variant of A. flavus, with a nicheability to thrive in the outbreak-hospital environment.