Pre-diabetes is a risk factor for the development of diabetes, not a disease in its own right. The prevalence increases with age and reaches nearly 50% of those aged over 75 years in the USA. While lifestyle modification and treatment are likely to benefit those with many years of life ahead of them, they are unlikely to benefit patients with a limited life expectancy. Despite this, some very elderly patients in the UK and elsewhere are being labelled as pre-diabetic. While ideal practice would be to carefully consider the impact of any potentially abnormal blood test before it is taken, this is not always possible in routine practice. In this paper, we discuss a pragmatic, ethical approach for clinicians managing pre-diabetic blood tests in very elderly patients. We argue that a ‘see-saw’ model of paternalism should be used in deciding which patients to inform that they can be labelled as pre-diabetic. Those patients that may benefit from the label should be informed, and those that will not, should not. Where the benefits/drawbacks are unclear, the result and its potential significance should be discussed in depth with the individual patient. We do not advocate withholding information from any patient. Instead we suggest clinicians use individual patient circumstances to contextualise the relevance of pre-diabetes to the patient and consider the benefits and drawbacks before informing them. This approach has the potential to be used for other pre-conditions and risk factors in addition to pre-diabetes.