Digital PCR (dPCR) is gaining popularity as a DNA mutation quantification method for clinical specimens. Fragmentation prior to dPCR is required for non-fragmented genomic DNA samples; however, the effect of fragmentation on DNA analysis has not been well-studied. Here we evaluated three fragmentation methods for their effects on dPCR point mutation assay performance. Wild-type (WT) human genomic DNA was fragmented by heating, restriction digestion, or acoustic shearing using a Covaris focused-ultrasonicator. dPCR was then used to determine the limit of blank (LoB) by quantifying observed WT and mutant allele counts of the proto-oncogenes KRAS and BRAF in the WT DNA sample. DNA fragmentation by heating to 95°C, while the simplest and least expensive method, produced a high background mutation frequency for certain KRAS mutations relative to the other methods. This was due to heat-induced mutations, specifically affecting dPCR assays designed to interrogate guanine to adenine (G>A) mutations. Moreover, heat-induced fragmentation overestimated gene copy number, potentially due to denaturation and partition of single-stranded DNA into different droplets. Covaris acoustic shearing and restriction enzyme digestion showed similar LoBs and gene copy number estimates to one another. It should be noted that moderate heating, commonly used in genomic DNA extraction protocols, did not significantly increase observed KRAS mutation counts.