Genetic characterization of the base population of a breeding program is useful in monitoring changes in the genetic structure that might occur over generations. By using microsatellite DNA markers, we have analyzed the genetic population structure of the Jamuna, Padma and Halda river samples of Catla catla maintained in a Brood bank of the Department of Fisheries, Bangladesh. We detected a total of 27 alleles at five loci in 90 individuals with an average of 5.4 alleles per locus. The mean allelic richness of the Jamuna, Padma and the Halda population of catla were 4.800, 5.200, and 4.400, respectively. The average observed (Ho) and expected (He) heterozygosity values for the Jamuna, Padma and the Halda rivers were 0.720, 0.720 and 0.667 and 0.663, 0.678 and 0.643, respectively. The FIS values ranged from -0.341 to 0.121 and found to be nonsignificant while the pairwise FST values ranged from 0.0142 to 0.0507 and found to be significant. The genetic variations in the Jamuna and the Padma populations are comparable and slightly higher than that of the Halda population. Our study detected an excess of heterozygosity indicating recent genetic bottlenecks in the brood bank founder stocks of C. catla.