Barriers to reproduction between species prevent hybridization and gene flow between heterospecific populations. In this dissertation I utilize two different, closely related pairs of Drosophilia fruit fly species to study the origin and maintenance of behavioral isolation, which is a key reproductive barrier between recently diverged species. First, I investigate the strength of behavioral isolation between D. occidentalis and D. suboccidentalis. I find that behavioral isolation in this system is asymmetric: crosses between D. suboccidentalis females and D. occidentalis males display stronger behavioral isolation than crosses between D. occidentalis females and D. suboccidentalis males. There is no evidence for postzygotic isolation between these species. Second, I determine if selection against hybridization between species can initiate divergence within an interacting species. If the hybrids generated between two species are unfit relative to pure species offspring, selection can increase the strength of premating discrimination against heterospecific mates in a process known as reinforcement. Furthermore, it has been proposed that reinforcing selection between species in sympatry can incidentally lead to strong behavioral isolation between sympatric and allopatric populations within an interacting species in a process known as cascade reinforcement. D. subquinaria females that co-occur with D. recens in nature display enhanced behavioral isolation not only against D. recens males, and they also display enhanced behavioral isolation against conspecific males derived from allopatric populations. To study how these two forms of behavioral isolation evolved in D. subquinaria, I used an experimental evolution approach in which I imposed reinforcing selection on non-choosy allopatric D. subquinaria females. I tracked the evolution of premating discrimination in these D. subquinaria females over time against D. recens males as well as against ancestral conspecific males. I find that reinforcing selection can generate stronger levels of behavioral isolation between species, both when hybrid fitness is 0 and when hybrid fitness is merely reduced relative to pure species offspring. However, I do not find evidence that D. subquinaria females developed enhanced behavioral isolation against ancestral conspecific males, implying that reinforcing selection alone may be insufficient to generate the patterns of cascade reinforcement observed in natural populations of this species.