The clinical chemistry of diabetes care is unique in Endocrinology, because many of the commonly performed tests, such as assays for glucose, ketones and hemoglobin A1c, are done by the patient at home or by the nurse at the bedside or clinic. Hence, some may assume that these tests are accurate and precise. However, measurement of these substances is often problematic. While laboratory-based glucose analyzers are generally accurate, pre-analytic variables may introduce confusion. Hand held glucose meters are in widespread use and a number of variables are assumed regarding their use, such as hematocrit, oxygen tension and pH. Urine ketone assays may be confusing during conditions altering the mitochondrial redox potential and they are prone to interference from medications. Many different assays for HbA1c are in use today and the various assay types may each be prone to interference from abnormal hemoglobins, iron deficiency states and medication use. International efforts are underway to standardize HbA1c measurements. Insulin assays are being more commonly used as the prevalence of type 2 diabetes and the metabolic syndrome increases. Unfortunately, insulin assays are not standardized, making comparisons of results from different laboratories difficult. Finally, the assessment of diabetes-associated autoantibodies is becoming clinically useful as our understanding of the immunologic basis of diabetes develops and as protocols for the prediction (and ultimately prevention) of the disease arise. While a physician's clinical sense should always guide the interpretation of laboratory values, an understanding of the basis of commonly ordered clinical tests is critical in the care of patients with diabetes.