We present a coupled hydrology and entomology model for the mechanistic simulation of local-scale response of malaria transmission to hydrological and climatological determinants in semiarid, desert fringe environments. The model is applied to the Sahel village of Banizoumbou, Niger, to predict interannual variability in malaria vector mosquito populations that lead to variations in malaria transmission. Using a high-resolution, small-scale distributed hydrology model that incorporates remotely sensed data for land cover and topography, we simulate the formation and persistence of the pools constituting the primary breeding habitat of Anopheles gambiae s.l. mosquitoes, the principal regional malaria vector mosquitoes. An agent-based mosquito population model is coupled to the distributed hydrology model, with aquatic-stage and adult-stage components. Through a dependence of aquatic-stage mosquito development and adult emergence on pool persistence, we model small-scale hydrology as a dominant control of mosquito abundance. For each individual adult mosquito, the model tracks attributes relevant to population dynamics and malaria transmission, which are updated as mosquitoes interact with their environment, humans, and animals. Weekly field observations were made in 2005 and 2006. A 16% increase in rainfall between the two years was accompanied by a 132% increase in mosquito abundance between 2005 and 2006. The model reproduces mosquito population variability at seasonal and interannual timescales and highlights individual pool persistence as a dominant control. Future developments of the presented model can be used in the evaluation of impacts of climate change on malaria, as well as the a priori evaluation of environmental management-based interventions.