The inappropriate use of self monitoring of glucose is wasteful of NHS resources and can cause psychological harm. Although a few patients find that self monitoring enables them to understand and take control of their diabetes, many people with diabetes are performing inaccurate or unnecessary tests. There is no convincing evidence that self monitoring improves glycaemic control, nor that blood testing is necessarily better than urine testing. It may be appropriate for some patients not to monitor their own glucose but to rely instead on regular laboratory estimations of glycaemic control. Glucose self monitoring should be performed only when it serves an identified purpose. It is widely assumed that glucose self monitoring, preferably of blood glucose concentrations, is desirable or even essential for everyone with diabetes. It is common for patients who have previously tested their urine, or have done no glucose monitoring at home, to be taught to measure their blood glucose when they are admitted to hospital. In the community too, patients are often encouraged to monitor their blood glucose, and newly diagnosed patients of all ages are usually taught to measure their blood glucose concentrations. Self monitoring can sometimes be useful, but evidence is mounting that its indiscriminate use is of questionable value. In 1995, Pounds 42.6 million was spent on home monitoring of glucose in the United Kingdom (Intercontinental Medical Statistics, personal communication). Is this enormous cost justified? Is blood testing necessarily better than urine testing? Is glucose self monitoring always necessary, or is it sometimes a waste of time and money? Are recommendations for self monitoring based on sound evidence?