The evolutionary histories of species are not measured directly, but estimated using genealogies inferred for particular loci. Individual loci can have discordant histories, but in general we expect to infer evolutionary histories more accurately as more of the genome is sampled. High Throughput Sequencing (HTS) is now providing opportunities to incorporate thousands of loci in ‘phylogenomic’ studies. Here, we used target enrichment to sequence c.3000 protein-coding exons in a group of Australian skink lizards (crown group age c.80 Ma). This method uses synthetic probes to ‘capture’ target exons that were identified in the transcriptomes of selected probe design (PD) samples. The target exons are then enriched in sample DNA libraries prior to performing HTS. Our main goal was to study the efficacy of enrichment of targeted loci at different levels of phylogenetic divergence from the PD species. In taxa sharing a common ancestor with PD samples up to c.20 Ma, we detected little reduction in efficacy, measured here as sequencing depth of coverage. However, at around 80 Myr divergence from the PD species, we observed an approximately two-fold reduction in efficacy. A secondary goal was to develop a workflow for analysing exon capture studies of phylogenetically diverse samples, while minimizing potential bias. Our approach assembles each exon in each sample separately, by first recruiting short sequencing reads having homology to the corresponding protein sequence. In sum, custom exon capture provides a complement to existing, more generic target capture methods and is a practical and robust option across low-moderate levels of phylogenetic divergence.