The two sibling cosmopolitan species, Drosophila melanogaster and D. simulans, are able to proliferate under very different climatic conditions. This has resulted in local adaptations, which are often arranged in latitudinal clines. Such clines are documented for body weight, wing and thorax length, sternopleural and abdominal bristle number, ovariole number and thoracic pigmentation. The overall magnitude of geographical differentiation is, however, much less in D. simulans than in D. melanogaster, and latitudinal clines are less pronounced. The fact that natural populations live under different climates raises the problem of interaction between temperature and phenotype. The reaction norms of morphometrical traits have been investigated as a function of growth temperature. The shapes of the response curves vary according to the investigated trait. They are generally curvilinear and can be described by calculating characteristic values after polynomial adjustments. For a given trait, the reaction norms of the two species are similar in their shape, although some significant differences may be observed. Within each species, significant differences are also observed between geographic populations: reaction norms are not parallel and the divergence is better marked when more distant populations (e.g., temperate and tropical) are compared. It thus appears that besides mean trait value, phenotypic plasticity is also a target of natural selection. A specific analysis of wing shape variation according to growth temperature was also undertaken. Reaction norms with different shapes may be observed in various parts of the wing: the major effect is found between the basis and the tip of the wing, but in a similar way in the two species. By contrast, some ratios, called wing indices by taxonomists, may exhibit completely different reaction norms in the two species. For a single developmental temperature (25 degrees C) the phenotypic variability of morphometrical traits is generally similar in the two species, and also the genetic variability, estimated by the intraclass correlation. A difference exists, however, for the ovariole number which is less variable in D. simulans. Variance parameters may vary according to growth temperature, and a detailed analysis was made on wing dimensions. An increase of environmental variability at extreme, heat or cold temperatures, has been found in both species. Opposite trends were, however, observed for the genetic variability: a maximum heritability in D. simulans at middle temperatures, corresponding to a minimum heritability in D. melanogaster. Whether such a difference exists for other traits and in other populations deserves further investigations. In conclusion, morphometrical analyses reveal a large amount of significant differences which may be related to speciation and to the divergence of ecological niches. Within each species, numerous geographic variations are also observed which, in most cases, reflect some kinds of climatic adaptation.