The production of tilapia in Brazil has shown continuous growth in the last decade and the development that has been observed recently is due to a transition from production models in floating cages in hydroelectric reservoirs to more closed systems, using excavated ponds. This process is also reinforced by the arrival of large animal protein companies focusing in the tilapia market, running an efficient and well-structured production chain that depends on good quality raw material to offer tilapia fillets to consumers in urban centers in Brazil and for export. In this study, eight production sites which production of tilapias had off flavor problems, pointed by consumers, were tracked. These specific production sites were studied in detailed environmental sampling of its limnology and planktonology. Analysis were carried out, seeking to relate cyanobacterial populations to the occurrence and persistence of off flavor, as indicated by sensorial analysis performed by a permanent, professional panel. Cyanobacterial populations were quantified through microscopic analysis and by fluorometric measurements "in vivo" using a multiparametric probe YSI EXO2 equipped with sensors for chlorophyll and phycocianins. These procedures were adopted seeking tools to monitor cyanobacterial populations in a timely manner for farm management decisions, that would allow the adoption of measures for mitigation of cyanobacterial populations. Other 41 earthern ponds were evaluated periodically with the multiparametric sonde and had fish collected for off flavor evaluation periodically along the grow out phase. The original set of eight ponds was selected for manipulations aiming to decrease cyanobacterial populations by increasing water flow, decreasing the cultivation density. The mitigating measures were to increase water exchange from 2-5% to 10% a day, and decreased feeding by 10-20%, associated with increased aeration. Emergency treatments were also tested to reduce the population of cyanobacteria, such as the application of lime and hydrogen peroxide in variable doses. Fluorometric measurements showed good correlation with microscopic quantifications, proving to be a useful tool for monitoring cyanobacterial populations. The cyanobacteria controlling procedures tested here proved to be efficient in decreasing the densities of cyanobacteria, but not of algae. Off flavor was significantly reduced when cyanobacteria had its density decreased. A protocol of off flavor mitigation measures was created, which involves improving the architecture of the tanks for less depth, increasing water flow, and decreasing the cultivation density from 10 to 5 fish per square meter. Emergency chemical treatments are being tested to reduce the population of cyanobacteria, such as the application of lime and hydrogen peroxide, with positive results, but still without a definitive protocol for their use although it appeared to be efficient both in acute cyanobacterial blooms and for the maintenance of low cyanobacterial densities. This study shows that tools available for cyanobacterial monitoring and mitigation can be effective in commercial tilapia production systems, and result in situations where off flavor can be eradicated or kept at insignificant levels avoiding losses for industry and