Introduction: Brazil is among the world’s largest producers of Nile tilapia (Oreochromis niloticus), a species increasingly challenged by a complex of cocirculating pathogens that threatens production sustainability.

Methods: We conducted a structured epidemiological surveillance program across the three main tilapia-producing macro-regions of Brazil (Southeast, Central-West, and Northeast) between 2023 and 2025. A total of 26,057 diagnostic samples, generating 68,514 pathogen-specific tests, were analyzed for Aeromonas spp., Edwardsiella spp., Flavobacterium spp., Francisella orientalis, Infectious Spleen and Kidney Necrosis Virus (ISKNV), Lactococcus spp., and Streptococcus spp. using standardized bacteriological, real-time PCR, and MALDI-TOF MS protocols, and antimicrobial susceptibility (minimum inhibitory concentration, MIC) was determined for four antimicrobials. Generalized linear models (with a farm-clustered mixed-model sensitivity analysis), time-series analysis (ADF stationarity, ARIMA), co-detection network analysis, and Gamma-log MIC modelling were applied.

Results: The cumulative overall detection positivity rate (DPR) was 18.1%, with Flavobacterium spp. showing the highest cumulative positivity (21.6%) on the smallest testing base, and Lactococcus spp. (18.4%), ISKNV (23.3%), and Streptococcus spp. (15.2%) representing the dominant, most widely sampled agents; S. agalactiae serotype Ia was not detected by serotype-specific qPCR. Detection was structured principally by macro-region and pathogen-specific temporal dynamics, whereas host developmental stage exerted a heterogeneous, generally weak influence. Five of the seven pathogens showed stationary, endemically compatible dynamics; Streptococcus spp. was non-stationary, consistent with active geographic expansion into the Northeast, whereas non-stationarity in Flavobacterium spp. reflected sampling instability rather than directional change. Quarterly aggregation minimized forecast error among the frequencies evaluated. Co-detection was predominantly independent, with only a weak module linking the three dominant agents. Antimicrobial resistance was structured by pathogen and antibiotic, with significant pathogen-specific interactions, and the fry stage showed higher MICs.

Discussion: These findings reframe the historically Streptococcus-centered view of Brazilian tilapia health and, within the constraints of a structured but non-probabilistic design, provide a descriptive, data-driven baseline to inform regionalized early-warning systems and rational antimicrobial stewardship in Brazilian tilapia aquaculture.