High-copy transposons have been effectively exploited as mutagens in a variety of organisms. However, their utility for phenotype-driven forward genetics has been hampered by the difficulty of identifying the specific insertions responsible for phenotypes of interest. We describe a new method that can substantially increase the throughput of linking a disrupted gene to a known phenotype in high-copy Mutator (Mu) transposon lines in maize. The approach uses the Illumina platform to obtain sequences flanking Mu elements in pooled, bar-coded DNA samples. Insertion sites are compared among individuals of suitable genotype to identify those that are linked to the mutation of interest. DNA is prepared for sequencing by mechanical shearing, adapter ligation, and selection of DNA fragments harboring Mu flanking sequences by hybridization to a biotinylated oligonucleotide corresponding to the Mu terminal inverted repeat. This method yields dense clusters of sequence reads that tile approximately 400 bp flanking each side of each heritable insertion. The utility of the approach is demonstrated by identifying the causal insertions in four genes whose disruption blocks chloroplast biogenesis at various steps: thylakoid protein targeting (cpSecE), chloroplast gene expression (polynucleotide phosphorylase and PTAC12), and prosthetic group attachment (HCF208/CCB2). This method adds to the tools available for phenotype-driven Mu tagging in maize, and could be adapted for use with other high-copy transposons. A by-product of the approach is the identification of numerous heritable insertions that are unrelated to the targeted phenotype, which can contribute to community insertion resources.