A laboratory investigation into the factors controlling migration in the African armyworm moth, Spodoptera exempta (Walker) was carried out using a tethered-flight technique. The methods used to culture S. exempta in the laboratory are described and evaluated., Larvae could not be satisfactorily reared on artificial diet, and a successful culture was only established when larvae were reared on freshly-cut maize leaves. A series of experiments were carried out to determine the effects of environmental conditions acting on the larvae on the flight performance of the resulting moths. Depriving larvae of food, or feeding them on leaves from, water-stressed maize plants, did not have any significant effects on the flight performance of female moths. However, when larvae were reared at high densities, a significant increase in the proportion of moths defined as "migrant" on an arbitrary criterion was observed. There was no significant difference between the wing-loadings of "migrant" and "non-migrant" moths. Observations made during the above experiments suggested that ä signifcant genetic component contributed to the determination of flight capacity. Selection experiments and the calculation of heritability estimates confirmed these expectations, and suggested that flight capacity was a polygenic character. Flight_ capacity of "migrant" moths generally declined with age. Feeding and mating did not have any , significant effects on the flight capacity of four-day old moths. No direct association could be found between flight capacity and rate of ovarian development. An hypothesis for the control of migration in S. exempta in East Africa is described. Differential selection is seen as operating between the wet and the dry seasons, with "flight"' genotypes selected for in the wet season when the potential habitat range of S. exempta expands, and "nonflight" genotypes being selected for during the dry season when habitats are isolated and topographically restricted.