1. The paper described an attempt to see whether differences in co-adaptation between populations ofDrosophila subobscuraare related to the distance between them. The mean and the variance of body-size, development time and survival were recorded on parent populations and the F1and F2of various crosses to test for heterosis in the F1and decline in performance or greater variance in the F2, which might indicate the break-up of co-adapted gene arrays. Comparisons were carried out at different temperatures and on a variety of larval diets, especially sub-optimal ones in which the larvae were grown on synthetic media. A large number of wild flies were caught at sites separated by about 10 miles along a transect of southern Scotland; these comprised one series of comparisons. For more distant crosses flies were caught at sites in southern England, Denmark, Switzerland and Israel.2. There were well-defined differences in body-size, and, to a lesser degree, development time between populations from more widely separated localities and these showed evidence of a cline, northern populations having larger body-size. The difference in size between the Scottish and Isreal populations is about 20%.3. There was no evidence of differences in co-adaptation between populations even in crosses between populations from sites as far apart as Scotland and Israel. The F1's were always close to the mid-parent values and there was no evidence of breakdown in the F2nor of increased variability.4. There was hardly any evidence of gene-environment interaction either with respect to different diets or to different temperatures.5. Records of body-size on flies caught in the wild showed that they are extremely variable, indicating great variation in larval nutrition. Under natural condition stability of growth in body-size is conspicuously lacking in this species.6. An additional test of co-adaptation was based on the between-family variance of abdominal bristle number of intra- and inter-population matings in the two most widely separated populations. There was no evidence of greater variance in the inter-population series.7. To test for possible differences in breeding structure, the response to inbreeding was determined for two widely separated populations ofD. subobscuraand a long-established cage population ofD. melanogaster, on an unrestricted larval diet and also on several different kinds of sub-optimal diets. There was little or no sign of consistent differences between the species in their response to inbreeding.8. This test revealed differences between the two species in their minimum requirements for particular nutrients.subobscurais less able thanmelanogasterto withstand lower levels of protein and survival is particularly reduced. On the other hand,melanogasterhas a considerably higher requirement for choline. Where there are apparent differences between the species in the average effect of inbreeding, the inbreeding effect is greater on the relatively more sub-optimal diet.9. Comparison of the performance of the immediate descendants of wild flies with those derived from the same site, but kept in the laboratory for some twenty generations, failed to show any differences on several different diets and so there was no evidence that adaptation to laboratory conditions was important.10. The lack of evidence for co-adaptation apparently conflicts with what has been claimed for other species. Such differences are discussed.