Sequence capture is a widespread and cost effective approach to resequence specific genomic regions on large sets of individuals. To date, this approach has mainly been used to target orthologous loci within the same species ('homologous' capture). Our aim was to test the efficiency of this method across closely or distantly related species ('heterologous' capture). Specifically, we focused on the MHC locus, which is one of the most dynamic regions of mammalian genomes, and a model to study host-pathogen coevolution. A 385K solid phase array (NimbleGen) encompassing the MHC was designed using the Sus scrofa Hp1a.0 haplotype as a reference (http://vega.sanger.ac.uk/index.html). Individuals from 9 species of Suidae (n=69) and 2 species of Tayassuidae (n=19) were selected for the study. Sets of 12 uniquely indexed sequencing libraries were prepared using Illumina kits, pooled, enriched, and sequenced on a HiSeq platform. In Suidae, the average values of coverage and specificity (% on target sequences) were respectively 86x and 32%. Results varied significantly across libraries, e.g. decreasing the number of cycles during PCR steps reduced the percentage of duplicate reads, but without improving specificity. On the other hand, in Tayassuidae both the coverage and the specificity showed lower efficiency. We can conclude that heterologous capture is a reliable technique at the level of Suidae, while in the case of Tayassuidae further optimization will be required to achieve better results. Other analyses are ongoing to test if this method could be suitable to study structural variation.