Circulating tumor DNA is being extensively investigated as a clinically relevant cancer marker. KRAS mutations are present in 40% of colorectal tumors and monitoring the mutational status together with the level of mutated DNA is of great interest. The measurement of DNA from plasma or serum, however, presents a number of challenges that require attention. The amount of DNA is low and highly fragmented and analyses need to be optimized accordingly. KRAS ARMS-qPCR assays with amplicon lengths of 120 and 85 base pairs, respectively, were compared using positive control material (PCR fragments) and plasma samples from 46 colorectal cancer patients known to harbor a tumor KRAS mutation. KRAS mutated DNA was detected in significantly more clinical samples using the short amplicon assays compared to the long amplicon assays (74% vs. 61%, p=0.03). The level of mutated DNA in plasma was on average three times higher using short amplicon assays. Our results reflect the importance of minimizing the assay length when analyzing highly fragmented DNA, especially if these analyses are to be used for treatment monitoring and relapse detection.