Since the ancient discovery of the 'sweet odor' in human breath gas, pursuits of the breath analysis-based disease diagnostics have never stopped. Actually, the 'smell' of the breath, as one of three key disease diagnostic techniques, has been used in Eastern-Medicine for more than three thousand years. With advancement of measuring technologies in sensitivity and selectivity, more specific breath gas species have been identified and established as a biomarker of a particular disease. Acetone is one of the breath gases and its concentration in exhaled breath can now be determined with high accuracy using various techniques and methods. With the worldwide prevalence of diabetes that is typically diagnosed through blood testing, human desire to achieve non-blood based diabetic diagnostics and monitoring has never been quenched. Questions, such as is breath acetone a biomarker of diabetes and how is the breath acetone related to the blood glucose (BG) level (the golden criterion currently used in clinic for diabetes diagnostic, monitoring, and management), remain to be answered. A majority of current research efforts in breath acetone measurements and its technology developments focus on addressing the first question. The effort to tackle the second question has begun recently. The earliest breath acetone measurement in clearly defined diabetic patients was reported more than 60 years ago. For more than a half-century, as reviewed in this paper, there have been more than 41 independent studies of breath acetone using various techniques and methods, and more than 3211 human subjects, including 1581 healthy people, 242 Type 1 diabetic patients, 384 Type 2 diabetic patients, 174 unspecified diabetic patients, and 830 non-diabetic patients or healthy subjects who are under various physiological conditions, have been used in the studies. The results of the breath acetone measurements collected in this review support that many conditions might cause changes to breath acetone concentrations; however, the results from the six independent studies using clearly-defined Type 1 and Type 2 diabetic patients unanimously support that an elevated mean breath acetone concentration exists in Type 1 diabetes. Note that there is some overlap between the ranges of breath acetone concentration in individual T1D patients and healthy subjects; this reminds one to be careful when using an acetone breath test on T1D diagnostics. Comparatively, it is too early to draw a general conclusion on the relationship between a breath acetone level and a BG level from the very limited data in the literature.