Understanding factors influencing sustained use of long-lasting insecticide-treated nets (LLIN) in areas of declining malaria transmission is critical to sustaining control and may facilitate elimination. From 2008 to 2013, 655 households in Choma District, Zambia, were randomly selected and residents were administered a questionnaire and malaria rapid diagnostic test. Mosquitoes were collected concurrently by light trap. In a multilevel model, children and adolescents of 5-17 years of age were 55% less likely to sleep under LLIN than adults (odds ratio [OR] = 0.45; 95% confidence interval [CI] = 0.35, 0.58). LLIN use was 80% higher during the rainy season (OR = 1.8; CI = 1.5, 2.2) and residents of households with three or more nets were over twice as likely to use a LLIN (OR = 2.1; CI = 1.4, 3.1). For every increase in 0.5 km from the nearest health center, the odds of LLIN use decreased 9% (OR = 0.9; CI = 0.88, 0.98). In a second multilevel model, the odds of LLIN use were more than twice high if more than five mosquitoes (anopheline and culicine) were captured in the house compared with households with no mosquitoes captured (OR = 2.1; CI = 1.1, 3.9). LLIN use can be sustained in low-transmission settings with continued education and distributions, and may be partially driven by the presence of nuisance mosquitoes.