Productivity of defective rainwater harvesting structures (RWHS) and other peridomestic habitats for dengue vector was assessed in a coastal town in Tamil Nadu, southern India, where dengue cases were reported. Of 31,709 houses, 792, 790, and 759 were surveyed during southwest monsoon, northeast monsoon, and summer season, and RWHS were found in 651, 638, and 544 houses, respectively. Of these RWHS, 23.3, 34.6 and 14.2% had defects; 20.7, 30.9, and 11.8% were holding water; and 6.5, 11.9, and 5.7% supported dengue vectors. Six types of RWHS, namely, open percolation pit, covered percolation pit, sealed percolation pit, rechargeable trench with bore, and structures connected to well in use or disuse were found. Number of female pupae of Aedes, i.e., pupal productivity obtained from RWHS ranged from 114 in summer to 1,174 in northeast monsoon, and open percolation pit contributed maximum (11.5%). The productivity of habitats other than RWHS varied from 635 in summer to 1,754 in northeast monsoon. Overall, the pupal productivity recorded from RWHS was 30.1% and in other habitats was 69.9%. A hierarchical cluster analysis showed three clusters of 23 types of habitats, which differed significantly in pupal production (F = 426.4, P < 0.05). Cluster I consists of 15 habitats, namely, disused well, RWHS-disused well, RWHS-used well, metal container, coconut shell, glass bottle, sealed percolation pit, used well, ornamental container, tree stump, defrost water collection tray, disposable cup, flower pot, broken toilet ware, and sun shade. Cluster II includes seven habitats, i.e., grinding stone, open percolation pit, covered percolation pit, cement tank, mud pot, plastic container, and rechargeable trench with bore, and cluster III includes automobile tire alone. Cluster II and III contributed to 80.1% of the total pupal production, whereas the remaining 19.9% by cluster I. The study showed that the defective RWHS, particularly open and covered percolation pits were found to be among the key containers, propagating Aedes population. The pupae per person obtained during northeast monsoon in different houses varied between 0.077 and 2.839 (average 0.864). House and Breteau Indices were relatively higher during northeast monsoon, whereas the Container Index was higher in southwest monsoon. In view of risk of dengue vectors propagation, the need for source reduction involving community and prioritizing control measures toward the highly productive water-holdings is discussed.