Understanding how reproductive isolation evolves and is maintained between species is central to speciation biology, yet the relative contributions of different isolating barriers and the genetic architecture of the traits that cause them remain understudied, especially for behavioral and polygenic traits. Here I address these questions using the Hawaiian swordtail crickets of the genus Laupala, a rapid radiation of 38 morphologically cryptic species that differ conspicuously in the pulse rate of the male mating song. Through no-choice mating trials, female phonotaxis assays, cuticular hydrocarbon profiling, and field surveys of microhabitat use in a naturally co-occurring Big Island species pair, I show that sexual isolation is nearly complete and sufficient to maintain species boundaries in sympatry despite weak habitat isolation. Using an F2 intercross mapping population, I identify two additive QTL underlying pulse rate divergence in a Maui species pair, both of which overlap with QTL previously mapped in independently evolving Big Island species pairs, providing evidence that genomic parallelism in song evolution is a genus-wide phenomenon. Gene Ontology enrichment analysis revealed an overrepresentation of neural signaling and rhythmic behavior genes within one QTL, with candidate genes like tipE, takeout, Ih, and Ace found in both QTL regions . Finally, I report the first case of natural hybridization in the Laupala radiation, confirming hybrid identity through population genomic analyses, demonstrating a strong correlation between genomic ancestry and pulse rate phenotype in nature, showing a heterogenous landscape of differentiation across the chromosomes and identifying some important candidate genes. Together, these results illuminate how sexual trait divergence drives and maintains reproductive isolation at multiple levels, from individual mating decisions to genetic architecture to genomic differentiation at a natural contact zone.