DNA metabarcoding is a relatively new molecular tool for dietary analysis of wildlife populations for monitoring and conservation. Metabarcoding involves use of universal primers to amplify a region of, for example, mitochondrial DNA (mtDNA) that is highly conserved within species and variable among species (a “barcode”) extracted from an environmental sample. By utilizing massively parallel sequencing of DNA extracted from predator feces, different prey mtDNA barcodes can be sequenced simultaneously to reveal dietary patterns. Because predator DNA also is abundant in their scats, sequences known as “blockers” are sometimes employed to block the amplification of host mtDNA while allowing prey mtDNA to be amplified preferentially. Before sequencing, a library must be constructed for each sample. Library construction involves amplifying the mtDNA barcode region using universal primers and attaching unique identifiers or indexes to those amplification products. In this study, we designed and tested four library-build protocols on red fox (Vulpes vulpes) fecal samples. Two protocols used universal primers attached to sample-specific inline tags, one treated with a blocker and the other without a blocker. The other two protocols used the universal primer without inline tags, relying on a second step for index attachment; one was treated with a blocker and the other without a blocker. We tested our protocols with 3 mock fecal samples (mixtures of prey DNA of known content) and 38 red fox fecal samples. We evaluated the performance of the four protocols based on total read numbers, prey read numbers, and the number of species identified. All four protocols successfully identified some of the prey species from the fox scats. Amplification without inline tag and with blocker yielded the highest average number of prey reads and species identified. Thus, we conclude that this protocol was the most efficient of the four.