Populations of D. melanogaster and D. simulans from Melbourne, Vic., and Brisbane, Qld, were regularly sampled for resistance to desiccation. D. melanogaster was more resistant than D. simulans and females of each species were more resistant than males for both populations.The resistance observed for the D. melanogaster population from Melbourne and for the populations of both species from Brisbane remained fairly constant throughout a yearly cycle. However, the resistance of the D. simulans population from Melbourne was greatest during the warmest months and became progressively less as the temperature decreased. For this D. simulans population percentage mortality was correlated with mean temperature and temperature fluctuation of the Melbourne environment. No other correlation was found for any other population comparison.Laboratory populations, derived from founders of each area and kept at constant temperatures of 15, 20 and 25°C, provided results that agreed with the field data. The level of resistance of the D. simulans population from Melbourne appeared to be related to the environmental temperature of maintenance but the other populations maintained fairly uniform resistance levels over a number of generations.Diallel cross techniques indicate that the genetic architectures for resistance to desiccation of the Melbourne D. melanogaster population and of both species from Brisbane are similar, with additive, dominance and interaction effects being observed. Desiccation resistance for D. simulans from Melbourne is almost entirely additive, however. This result is discussed in relation to the field and laboratory data.