Hydropower is an essential part of the solution to climate change; however, large dams are questionable as a reliable strategy, since they modify the entire ecosystem downstream, affecting the livelihood of riverine communities. This is concerning in regions of high freshwater biodiversity, such as Brazil, which happens also to be the third largest hydropower producer. Whenever fish passages are not available, restocking actions have been applied in hydropower dams impacted river basins, mostly without adequate protocols or monitoring, and focusing on a quantitative, rather than a conservative perspective. One of the most power plants-impacted rivers of Brazil is the São Francisco River, among the largest river basins in Brazil, where restocking actions have been carried out for the past 30 years. Five species were investigated in terms of genetic structure using nuclear markers, three occurred as single populations, while two existed as metapopulations. Two highly impacted stretches of the river (submiddle and lower) need to invest in stocking actions, while the middle stretch should not carry out any such operations and become an Environmental Protection Area (EPA), due to the continuous abundance of migratory fish species and to the several tributaries that provide nutrients into the system. Among the four studies that compared hatchery-captive and wild broodstocks, losses in terms of the number of alleles were detected in all of them, varying between 20 and 50%. Also, the seminatural breeding strategy adopted in some restocking hatcheries, where breeders are hormonally induced and spawning occurs naturally in spawning tanks, does not guarantee equal parental contribution and only 15% of male breeders effectively contribute to the next generation. The lack of parental information in terms of age, sex, sperm quality, and genetic relatedness coefficient makes the challenge of breeding for restocking a task with an unpredictable success. Successful restocking programs rely on the quality of the information for the broodstock, on the knowledge of how wild populations are structured, on ecological information of life history of the wild conspecifics, on the determination of the ideal size for release, and finally, on the traceability of the restocked individuals to measure the efficiency of such activity.