Of eight species of Drosophila (ananassae, funebris, melanogaster, nebulosa, pseudobscura, simulans, virilis, and willistoni), the number of progeny produced per culture vial increases with the number of female parents for D. melanogaster, D. simulans, D. willistoni, D. ananassae, and D. nebulosa; in contrast, the progeny yield is a constant that is independent of the number of parental females per vial for D. funebris, D. pseudoobscura, and D. virilis. Forall species, however, the per capita yield decreases more or less exponentially with the square—root of the number of parental females per vial. The outcome of mixed—species competition depends largely upon the two species concerned. Drosophila simulans seemingly benefits by increases in number of its own females; four or five D. simulans females yield more progeny in the presence of D. funebris, for example, than produces D. simulans. Drosophila simulans alsoproduces greater number of progeny per vial as the number of competing D. funebris females is increased. Drosophila melanogaster and D. simulans females competing in individual culturesyield mixed progeny; the two species share the resources of each vial. In contrast, the progeny of D. pseudoobscura and D. virilis females in individual mixed—culture vials tend to be of one species or the other many cultures containing these two types of females yield few or no flies of either species. For most species—pairs the logarithm of the ratio of the two species among offspring varies directly with that among parents (slope of the regression = 1.00). For three species—pairs, however, the slope of this regression is significantly less than 1.00. If the same relationship were to hold among freelybreeding populations of the two species, the species should establish stable equilibrium proportions. Analyses of the competing species by means of replacement series reveals that the members of individual pairs may enhance or interfere with each other's development, or do neither. Similarly, the competitive abilities of the species differ considerably. The ease with which each species uses different resources within the culture vials differs. Thus, intra— and interspecific competition differentially affect the per capita progeny production of conspecific and heterospecific females. Because the competition between conspecific females differs demonstrably from that between heterospecific ones, it appears that inter—specific frequency—dependent selection may reflect the relative proportions of resources that are used with contrasting ease by the competing species. If so, frequency—dependent selection is precisely the sort of evidence needed to demonstrate that two species are not competing identically for identical resources, the type of competition that is expected to lead to competitive exclusion.