The clinical implications of being able to accurately detect tumor-derived DNA in the circulation, termed circulating tumor DNA (ctDNA), could be enormous. Already, a plethora of clinical applications is under validation that include detection of minimal residual disease and predicting recurrence, monitoring response and resistance to treatment, identifying targets for therapies, and early detection. ctDNA is only a fraction of the total cell-free DNA (cfDNA) which confounds its detection and sometimes conceals its properties. To use ctDNA as a cancer biomarker with confidence, we need to understand its nature. Its characteristics, including size, half-life, and amount, are critical for the development of tests for its detection and discrimination from the rest of the cfDNA. Technological advances have enabled the detection and quantification of individual fragments of cfDNA, which is pivotal for clinical applications. Understanding the causes, the source of and the mechanisms of release of ctDNA are important for the interpretation of test results. Despite the many advances in understanding the nature and biology of ctDNA, we do not yet have a clear appreciation of the processes that govern its presence and levels in the circulation. ctDNA is not detectable in the blood of every cancer patient, and there is not a directly proportional relationship to tumor type, size, or stage. It is not clear if the lack of correlation with these specific clinical parameters is strictly due to technical or biological challenges. Better understanding of the pathophysiology of ctDNA is therefore important for the improvement of clinical applications and interpretation of their results.