Wing dimorphism occurs commonly among many species of insects. This di
morphism is hypothesized to be maintained by a trade-off between the c
osts of being macropterous (winged, flight capable) and the long-term
benefits of migration in a heterogeneous environment. In this article
this hypothesis is investigated with a simulation model consisting of
an environment composed of discrete patches that persist for a fixed p
eriod and a generalized life-history pattern. The analysis shows that
a dominant brachyptery (rudimentary wings, flight incapable) allele ca
n readily spread in a monomorphically macropterous population and that
at equilibrium genetic variability is maintained. The invasion of the
brachyptery allele depresses population size but enables the populati
on to spread into regions in which the monomorphically macropterous po
pulation cannot persist. Further analysis of the model shows that patc
h persistence time (T), the cost of being macropterous (c), the propor
tion of migrants (m), and the probability of locating another patch (s
) are all important in determining the frequency of macroptery but tha
t T and c are the most important. The model predictions are compared w
ith data on planthoppers: an excellent fit is obtained. This fit is ro
bust to parameter values for persistence times greater than five gener
ations. The implications of this model with respect to the evolution o
f migration and dimorphisms are discussed.