Birdsong is a sexually selected trait that could play an important evolutionary role when related taxa come into secondary contact. Many songbird species however learn their songs through copying one or more tutors, which complicates the evolutionary outcome of such contact. Two subspecies of a presumed vocal learner, the grey-breasted wood-wren (Henicorhina leucophrys), replace each other altitudinally across the western slope of the Ecuadorian Andes. These subspecies are morphologically very similar, but show striking differences in their song. We examined variation in acoustic traits and genetic composition across the altitudinal range covered by both subspecies and between two allopatric populations. The acoustic boundary between the subspecies was found to be highly abrupt across a narrow elevational range with virtually no evidence of song convergence. Mixed singing and use of hetero-subspecific song occurred in the contact zone and was biased towards the use of leucophrys song types. Hetero-subspecific song copying by hilaris and not by leucophrys reflected a previously found asymmetric pattern of response to song playback. Using AFLP markers, we detected hybridization in the contact zone and asymmetric introgression in parapatric populations, with more leucophrys alleles present in hilaris populations than vice versa. This pattern may be a trail of introgression due to upslope displacement of leucophrys by hilaris. Our data suggest that song learning may impact speciation and hybridization in contrasting ways at different spatial scales: while learning may speed up population divergence in songs, thereby enhancing assortative mating and reducing gene flow, it may at a local level also lead to the copying of heterospecific songs, therefore allowing some level of hybridization and introgression.,Song-measurements_Dryadfile containing all song measurements. For each individual bird we recorded several song types. We obtained 22 acoustic measurements for each song type. We analysed three renditions of the same song type per bird and data presented here is averaged over these song types.,