Insecticide resistance is a major impediment to the control of vectors and pests of public health importance and is a strongly selected trait capable of rapid spread, sometimes even between closely related species. Elucidating the mechanisms generating insecticide resistance in mosquito vectors of disease, and understanding the spread of resistance within and between populations and species are vital for the development of robust resistance management strategies. Here, we studied the mechanisms of resistance in two sympatric members of the Anopheles gambiae species complex-the major vector of malaria in sub-Saharan Africa-to understand how resistance has developed and spread in eastern Uganda, a region with some of the highest levels of malaria. In eastern Uganda, where the mosquitoes Anopheles arabiensis and An. gambiae can be found sympatrically, low levels of hybrids (0.4 %) occur, offering a route for introgression of adaptively important variants between species. In independent microarray studies of insecticide resistance, Gste4, an insect-specific glutathione S-transferase, was among the most significantly up-regulated genes in both species. To test the hypothesis of interspecific introgression, we sequenced 2.3 kbp encompassing Gste4. Whilst this detailed sequencing ruled out introgression, we detected strong positive selection acting on Gste4. However, these sequences, followed by haplotype-specific qPCR, showed that the apparent up-regulation in An. arabiensis is a result of allelic variation across the microarray probe binding sites which artefactually elevates the gene expression signal. Thus, face-value acceptance of microarray data can be misleading and it is advisable to conduct a more detailed investigation of the causes and nature of such signal. The identification of positive selection acting on this locus led us to functionally express and characterise allelic variants of GSTE4. Although the in vitro data do not support a direct role for GSTE4 in metabolism, they do support a role for this enzyme in insecticide sequestration. Thus, the demonstration of a role for an up-regulated gene in metabolic resistance to insecticides should not be limited to simply whether it can metabolise insecticide; such a strict criterion would argue against the involvement of GSTE4 despite the weight of evidence to the contrary.