The potential benefits and risks of genetically engineered gene-drive systems for replacing wild pest strains with more benign strains must be assessed prior to any field releases. We develop a computer simulation model to assess the feasibility of using engineered underdominance constructs to drive transgenes into age- and spatially structured mosquito populations. Our practical criterion for success is the achievement of a transgene frequency of at least 0.80 within 3 years of release. The impacts of a number of parameters that may affect the success of gene-drive, such as the release area, release age, density-dependent larval survival, fitness cost of the engineered genes, and migration probability of adults, are examined. Results show that patchy release generally requires the release of fewer engineered insects to achieve success than central release. When the fitness cost is very low, central release covering 25% of the total area can be more effective than a completely uniform release over the whole area. This study demonstrates that to determine the best method of spatial release, and the total number of engineered insects that must be released, it is important to take into account the age and sex of the released insects and spatial structure of the population.