Determining the genomic locations of transposable elements is a common experimental goal. When mapping large collections of transposon insertions, individualized amplification and sequencing is both time consuming and costly. We describe an approach in which large numbers of insertion lines can be simultaneously mapped in a single DNA sequencing reaction by using digital error-correcting codes to encode line identity in a unique set of barcoded pools. N EXT-generation sequencing (NGS) technologies have greatly reduced the cost of DNA sequence analysis through the parallel sequencing of many short fragments. However, many applications, including molecular cloning and mutational analysis continue to rely on conventional capillary electrophoresis Sanger sequencing methods, as these are well-suited to sequencing individual fragments. Thus, one challenge in using NGS technologies in such applications lies in preserving sample identity while sequencing many samples simultaneously. This challenge can be addressed by encoding sample identity through either DNA barcoding or directed pooling (Mazurkiewicz et al. 2006;Erlich et al. 2009;Goodman et al. 2009;Prabhu and Pe'er 2009).Transposable elements represent powerful tools for manipulating the genomes of many model organisms (Bellen et al. 2011;Bire and Rouleux-Bonnin 2012). Thus, determining the genomic location of transposon insertion sites is a common experimental goal. Several methods, such as inverse PCR and splinkerette PCR, are used to amplify a short fragment of the genome directly adjacent to an insertion (Ochman et al. 1988;Devon et al. 1995). Subsequently, capillary electrophoresis Sanger sequencing is used to sequence each amplicon. As a result, all processing reactions must be performed independently on each sample, making the cost and labor associated with mapping collections of thousands of insertion lines significant. Several techniques have used NGS to map transposons in large populations of bacteria or yeast (Goodman et al. 2009;Uren et al. 2009;Iskow et al. 2010;Febrer et al. 2011). However, most of these approaches do not allow the insertion site to be associated with the identity of the original sample.While DNA barcoding can be used to encode sample identity prior to NGS, adding the barcode requires either individualized molecular manipulation of each sample or prior construction of a sequence-tagged transposon library (Mazurkiewicz et al. 2006;Hamady et al. 2008). As an alternative, pooling strategies can be used to encode sample identity, and several recent studies have reported strategies for efficient, error-resistant encoding in pooled DNA samples (Erlich et al. 2009;Goodman et al. 2009;Prabhu and Pe'er 2009). However, none of these approaches have been applied to the large genomes of multicellular eukaryotes, which present unique challenges due to repetitive sequences, increased sequence complexity, and an 100-to 1000-fold reduction in the ratio of transposon sequence to genome sequence.We have developed a method for mapping transposo...