Diabetes is a chronic disease affecting a significant portion of the human population, namely those above a certain age. The condition often progresses from a stage known as “pre-diabetes” when the glucose concentration in blood before any fasting in the morning is above a conventional limit of 100 mg/dL. The engagement is much smaller in the phase of pre-diabetes than during the full disease, so that it must be carefully controlled, and normally drugs are needed to avoid stepping to the full disease. Consequently, regular tests of the glucose concentration are mandatory to confirm the “pre” (or better) condition. Frequent home tests are obviously a heavy burden for patients. In scientific terms, this means that statistics related to the concentration trend over time are useful and necessary: they are obtained by regularly using commercial “strips” and their testers, with instructions on how to handle them. The present study found unexpected drawbacks that make such tests unreliable, irrespective of the patient performing the measurements correctly.