Displacement rates of tsetse affect performance of targets during vector control. Fly size, one of the indicators of population structure usually obtained from wing measurement, is among the determinants of displacement rates. Although recovery of tsetse in previous intervention areas has been widely reported, the population structure of tsetse that recover is rarely evaluated despite being associated with displacements rates. Previously, intervention trials had reduced tsetse densities by over 90% from >3 flies/trap/day to <1fly/trap/day on Big Chamaunga and Manga islands of Lake Victoria in western Kenya. In this study, we assessed the recovery in densities of Glossina fuscipes fuscipes on the two islands and evaluated the effects vector control might have on the population structure. A before and after intervention study was undertaken on four islands of Lake Victoria in western Kenya; Small and Big Chamaunga, Manga and Rusinga Islands, two of which tsetse control intervention had previously been undertaken. Three years after intervention average G. f. fuscipes catches in biconical traps were estimated on each island. Wing centroid size (CS) (a measurement of fly size) and shape, indicators of the population structure of flies from the four islands were compared using geometric morphometric analyses. CS and shape of available female but not male tsetse wings obtained before the intervention trial on Big and Small Chamaunga islands were compared with those from the same islands after the intervention trial. G. f. fuscipes apparent density on the previous intervention islands were>9 flies/trap/day. Irrespective of sex, wing shape did not isolate tsetse based on their islands of origin. The fly size from Big and Small Chamaunga did not differ significantly before intervention trials (P = 0.728). However, three years after the intervention flies from Big Chamaunga were significantly smaller than those from Small Chamaunga (P < 0.003). Further, there was an increase in the divergence of wing morphology between flies collected from Big Chamaunga and those from Small Chamaunga after tsetse control. In conclusion, even though populations are not isolated, vector control could influence the population structure of tsetse by exerting size and wing morphology differential selection pressures. Therefore, we recommend further studies to understand the mechanism behind this as it may guide future vector control strategies.