Background: The “Liquid Biopsy” has become a powerful tool for cancer research during the last decade. Circulating cell-free DNA (cfDNA) that originates from tumors has emerged as one of the most promising analytes. In contrast to plasma-derived cfDNA, only a few studies have investigated urinary cfDNA. One reason might be rapid degradation and hence inadequate concentrations for downstream analysis. In this study, we examined the stability of cfDNA in urine using different methods of preservation under various storage conditions. Methodology: To mimic patient samples, a pool of healthy male and female urine donors was spiked with a synthetic cfDNA reference standard (fragment size 170 bp) containing the T790M mutation in the EGFR gene. Spiked samples were preserved with three different buffers and with no buffer over four different storage periods (0 h; 4 h; 12 h; 24 h) at room temperature vs. 4 °C. The preservatives used were Urinary Analyte Stabilizer (UAS, Novosanis, Wijnegem, Belgium), Urine Conditioning Buffer (UCB, Zymo, Freiburg, Germany) and a self-prepared buffer called “AlloU”. CfDNA was extracted using the QIAamp MinElute ccfDNA Mini Kit (Qiagen, Hilden, Germany). CfDNA concentration was measured using the Qubit™ 4 fluorometer (Thermo Fisher Scientific, Waltham, MA, USA). Droplet digital PCR (ddPCR) was used for detection and quantification of the T790M mutation. Results: Almost no spiked cfDNA was recoverable from samples with no preservation buffer and the T790M variant was not detectable in these samples. These findings indicate that cfDNA was degraded below the detection limit by urinary nucleases. Stabilizing buffers showed varying efficiency in preventing this degradation. The most effective stabilizing buffer under all storage conditions was the UAS, enabling adequate recovery of the T790M variant using ddPCR. Conclusion: From a technical point of view, stabilizing buffers and adequate storage conditions are a prerequisite for translation of urinary cfDNA diagnostics into clinical routine.