Background: Over the last two decades, there has been remarkable progress in malaria control in sub-Saharan Africa, due mainly to the massive deployment of long-lasting insecticidal nets and indoor residual spraying. Despite these gains, it is clear that in many situations, additional interventions are needed to further reduce malaria transmission. The World Health Organization (WHO) has promoted the Integrated Vector Management (IVM) approach through its Global Vector Control Response 2017-2030. However, prior roll-out of larval source management (LSM) as part of IVM, knowledge on ecology of larval aquatic habitats is required. Methods: Aquatic habitats colonized by immature Anopheles and culicines vectors were characterized at three sites of low, medium and high malaria transmission in Uganda from October 2011 to June 2015. Larval surveys were conducted along transects in each site and aquatic habitats described according to type and size. Immature Anopheles, culicines and pupae from the described habitats were sampled using standard dipping methods to determine larval and pupae densities. Larvae were identified as anopheline or culicine, and counted. Pupae were not identified further. Binary logistic regression analysis was used to identify factors associated with the presence of immature Anopheles and culicines in each site. Results: A total of 1205 larval aquatic habitats were surveyed and yielded a total of 17,028 anopheline larvae, 26,958 culicine larvae and 1189 pupae. Peaks in larval abundance occurred in all sites in March-May and August-October coinciding with the rainy seasons. Anopheles larvae were found in 52.4% (n = 251) of aquatic habitats in Tororo, a site of high transmission, 41.9% (n = 536) of habitats in Kanungu, a site with moderate malaria transmission, and 15.8% (n = 418) in Jinja, a site with low malaria transmission. The odds of finding larvae was highest in rice fields compared to pools in both Tororo (odds ratio, OR = 4.21, 95% CI 1.22-14.56, p = 0.02) and Kanungu (OR = 2.14, 95% CI 1.12-4.07, p = 0.02), while in Jinja the odd were highest in containers (OR = 4.55, 95% CI = 1.09-19.14, p = 0.03). In Kanungu, larvae were less likely to be found in containers compared to pools (OR = 0.26, 95% CI 0.09-0.66, p = 0.008) and river fringe (OR = 0.19, 95% CI 0.07-0.52, p = 0.001). Medium sized habitats were associated with high odds of finding larvae compared to small habitats (OR = 3.59, 95% CI 1.18-14.19, p = 0.039).