DNA barcoding has become a dependable way to assign taxonomic identifications to otherwise unknown DNA samples. Metabarcoding using next-generation sequencing (NGS) maintains this same caliber of taxonomic identification and can occur on a mass scale in pooled, or multiplexed, samples from a wide variety of sources when properly multiplexed. However, this methodology has several pitfalls, including inaccurate source sample assignment, multi-step workflows that leave room for human error, and high costs. To combat these issues, library preparation protocols have been established for the popular Illumina sequencing platform which reduce sequence assignment errors to levels equivalent to background noise, while also greatly reducing cost. An equivalent protocol has yet to be established for the Ion Torrent sequencing ecosystem. Here, we present MetaPlex, a library preparation workflow and post-processing toolkit for efficient and accurate COI metabarcoding on Ion Torrent sequencers. These methods significantly decrease the costs of multiplexed sample sequencing by nearly 9x when compared to commercially available kits through the reduction of reagents needed for library preparation. In addition, our workflow provides increased reliability through elimination of laboratory processes which are known to negatively bias NGS outputs and source-sample tracing, limiting the common pitfall known as index jumps to 0.17%. Finally, our accompanying bioinformatic toolkit increases user accessibility by providing an easy-to-use command line tool for processing and correcting for errors in our, and others’, dual-indexed reads.